Jump to content
TUFLOW Forum

ade

Inactive members
  • Content Count

    6
  • Joined

  • Last visited

Community Reputation

0 Neutral

About ade

  • Rank
    Member
  1. I have a problem with a model experiencing a consistently large 1D/2D mass error. The mass error only commences with outflow and then gradually increasing in magnitude (-9% for longer events). The 1D/2D mass error increases while inflow is greater than outflow. Once the outflow is larger ME starts to gradually diminsh and approach 0%. The 1D mass error is ~0%, the 2D remains within +/-1%, there is low fluctuation in change in volume (i.e. not jumping) and there is no instability. Any ideas? Background: The model is of a medium top of catchment approximately 580ha in size represented by a single 2D domain. Large pipes(>450mm) and channels have are represented as part of a 1D domain using ESTRY. Outflow is from the 2D domain and modelled as both a HQ boundary.
  2. I have a 1d/2d model consisting of two converging channels and an underground network (1D domain) within a ~300Ha urban catchment (2D domain). The model appears to become unstable once there is flow from the channel to the 2d domain. Upon inspection the z-pts check file reveals zc points have been raised in and around the problem location (and HX line) by up to 2m from the original z-pts layer. The Z switch has been used in the HX layer (errors produced without). The z-pts compare well with the 1D cross sections and should not be raised by such a large amount. Any ideas?
  3. Bill, will the new release be accompanied by updated documentation? You also mentioned a wiki during your presentations. How is this going? Ade
  4. Hi Sam You will find that if you are after the maximums q is taken at the maximum water surface, Z0 is taken as the maximum vXd over all timesteps. The two are not necessarily the same. It sounds like you are interested in Z0. Regarding the hazard catergories I know "Map Output Data Types == ZMW1 ZMW2" switches are specifically for Melbourne Water's preferred outputs. However I am unsure of the format, details are not documented. I suggest you play it safe and stick with Z0 and simply post-processed further, at least until the switches are fully documented. Hope this is of help. Ade
  5. ade

    1D/2D instability

    Hi Richard Most of those arrangements I have tried but was not sure which to settle on. Now i do. I still get instability but it takes some unknowns (for me) out. Thanks Richard. With any luck your suggestions, along with pm's, will hopefully be enough to sort it out. Ade
  6. Basically, I have a model with a gradually increasing mass error and eventually going unstable. I have tried everything I can think of, looked at the message files, check files, changed the arrangement, time steps and others with no success. I suspect it is to do with the 1D-2D connection at the upstream end of the channel, it is a messy arrangement. There is a roadway acting like a weir with flow along and across the downstream face, and, a transition from pipe to channel in the same location. Further, it is difficult to get a HX line at this point perpendicular to flow for all flows. I have done the best I can but suspect it could be improved. I have included an image below with ME grid in the background for a very extreme event (PMP) but the problem also occurs for lesser events. CD = green cross hatching XS = brown QT bc = magenta HX = orange (broken) (also obsured by CD perimeter) CN = light blue (also obscured by XS) pipe = blue (solid) channel = blue (broken),
×
×
  • Create New...