WinFlume Revision History
- 05/11/12 - Executable file stamp 1.06.0004
- Modified curve fit routine to allow head offset term K2 to be negative. For some flumes, this gives a better curve fit result.
- Added a msgbox prompt for auto-matching control width to approach channel width if the user makes the control equal to a copy of one of the other sections.
- Modified RescaleForm routine to do a better job of scaling large text boxes, like the text box used on the Flume Data Report / Design Report screen.
- Modified positioning of the Match Sill Width button on cross section editing form, so that it doesn't cover up the section label.
- 06/27/11 - Executable file stamp 1.06.0003
- Fixed a problem with design evaluation report for movable-crest flumes.
- 05/16/11 - Executable file stamp 1.06.0002
- Fixed bug that was causing error 438 when trying to print flume drawing.
- 02/28/11 - Executable file stamp 1.06.0001
- Added a feature to allow change in units while preserving numerical values of entered dimensions. This is useful when you have accidentally forgotten to set proper units before entering dimensions. Also added more visible units labels on bottom profile and section views to help us users more readily notice when units are improperly set.
- Changed text on "Match to Approach Channel Width" button to say instead, "Make Sill Span Channel Width". Many users in our training classes don't pick up very well on what this button does; maybe this alternate text will help.
- Provided more verbose suggestions on how to deal with issue of a converging transition that needs to be longer due to sidewall contraction. The displayed text now suggests that the ramp be constructed at a steeper slope than the slope that will be shown for the long sidewall transition.
- Also added text to the bottom of the design review report to further emphasize previous point, and give guidance on ensuring a smooth flow at the staff gage position (no offsets upstream from staff gage within distance of 1*H1_max).
- Added a new button on the bottom profile form that will attempt to automatically adjust the lengths of the key flume components to reasonable values for the design discharge range. This function will set the control (throat) length, the converging section length, and the approach channel length (distance from gage to start of converging transition).
- 04/24/09 - Executable file stamp 1.05.0033
- Fixed overflow error occurring with flumes having discharges with very large magnitude, either due to choice of units or size of flume.
- 10/08/08 - Executable file stamp 1.05.0032
- Fixed screen calibration form so that it does not
rescale to fit user's screen size. If user has an
unusual screen aspect ratio (e.g., a portrait screen) it will cause the blue box to scale differently than the font size, which defeats the design of the calibration process. This only affects on-screen appearance of wall gages.
- Fixed screen calibration form so that it does not rescale to fit user's screen size. If user has an
- 07/07/08 - Executable file stamp 1.05.0031
- Fixed bug that caused division-by-zero error when checking the throat length to base-width ratio of flumes with a V-shaped control and no raised sill.
- 04/09/08 - Executable file stamp 1.05.0030
- Fixed bug that allowed the measurement duration to be less than the measurement interval (due to a change of units on the duration field), which leads to a divide by zero error when computing the uncertainty of totalized flow.
- Made a fix to set the measurement interval to at least one second if it is zero. Some flume files I have seen have the measurement interval set to 0 seconds (I can't find a way for this to occur, but it apparently does). The fix corrects such a problem if it ever occurs.
- 8/13/07 - Executable file stamp 1.05.0029
- Fixed problem with the definition of the focal distance
for parabolic sections. The program actually utilizes
the distance from the focus to the directrix, which is
two times the focal distance. Just changed labels so
that it is correctly identified now. Also rebuilt setup
kit to update all DLL's because I noticed TABCTL32.OCX was not synchronized on my system, causing rating tables to appear backwards (fixed columns at right).
- Fixed problem with the definition of the focal distance for parabolic sections. The program actually utilizes the distance from the focus to the directrix, which is two times the focal distance. Just changed labels so that it is correctly identified now. Also rebuilt setup
- 5/18/06 - Executable file stamp 1.05.0028
- Problem with graph legend labels was still not completely fixed in version 1.05.0026. Now it is, and graphs are correct even when some of the data to be plotted have warning messages.
- 4/30/05 - Executable file stamp 1.05.0027
- Modified column headings on equation report and rating tables to eliminate most of the use of the term "error" and better describe differences between theoretical discharge computed by WinFlume's analysis and the fitted discharge obtained from the simplified rating equation obtained by curve fitting. Also replaced the term "error" with uncertainty in reports on the total measurement error (i.e., combined rating table and water level measurement uncertainties).
- 4/12/05 - Executable file stamp 1.05.0026
- Graph legend labels on the curve-fit equation plot have been backward all these years. Curve fits were so close that it apparently was never obvious. I reversed the legends so they are now correctly labeled. Data points are actual theoretical rating table results, while line shown is curve-fit equation.
- 1/5/05 - Executable file stamp 1.05.0025
- Modified the criteria used to generate warning messages
regarding the length-to-width ratio of flumes that are
solely width-contracted. The new criteria is much less
restrictive. It will issue a warning only when the
L/W ratio is less than 1 (previous limit was 2), and
the submergence ratio is above the minimum modular limit
associated with the throat shape (see Table 2.1,
Water Measurement with Flumes and Weirs). This latter check is entirely new and is based on Wahl et al. (2002). Finally, the value of W used in the check has been changed from the average of the crest width at the sill and at the h1 level to just the value at the sill. These changes were prompted by field reports of numerous width-contracted flumes having L/W between 1 and 2 that have been verified accurate by current meter measurements.
- Modified the criteria used to generate warning messages regarding the length-to-width ratio of flumes that are solely width-contracted. The new criteria is much less restrictive. It will issue a warning only when the L/W ratio is less than 1 (previous limit was 2), and the submergence ratio is above the minimum modular limit associated with the throat shape (see Table 2.1,
- 12/21/04 - Executable file stamp 1.05.0024
- Added new section to design review report that provides an estimate of the uncertainty of average flows or totalized flow volumes obtained from multiple discharge measurements. New data input from user needed for this feature is located on the Head Measurement tab of the flume data screen.
- Released Excel spreadsheets for computing flume construction quantities estimates. These are not a part of the WinFlume program itself, but are posted on the WinFlume home page.
- Fixed bug that was preventing the loading and conversion of tailwater lookup tables when an old .flm file is loaded into a newer version of WinFlume.
- 3/9/04 - Executable file stamp 1.05.0023
- Fixed problem with design evaluation of flumes with zero sill height, using the "raise and lower sill" method. Previous versions were not allowing sill heights less than 0.3 meters, although with some width contraction, such designs are certainly feasible.
- Fixed problem with program sometimes not prompting at the appropriate time to save design modifications (if user closes flume dimensions form using the Windows control in upper right corner of form, rather than using the program's File Close or File Exit menu options.
- Shortened up a few of the hydraulic warning messages so that they can be displayed completely in the brief design review area. Previously they ran off the right side of the screen and could not viewed without scrolling.
- 12/30/02 - Executable file stamp 1.05.0022
- Fixed problem with overflow error when column increment for a ditchrider's rating table is exceedingly small, producing a table that contains more than 32767 entries.
- 12/17/02 - Executable file stamp 1.05.0021
- Fixed problem with division by zero errors when design discharge is greater than the capacity of a circular control section.
- 11/1/02 - Executable file stamp 1.05.0020
- Fixed problem with flumes having no sill height causing an error in QtoH routine if approach channel and control section have identical shape and dimensions. Note that this never could be obtained as a final design because there is no contraction to cause critical depth, but this combination of conditions can occur inadvertently while a user is developing a design or using the flume wizard to enter a design.
- 5/21/02 - Executable file stamp 1.05.0019
- Fixed problems with taking square roots of slightly negative numbers in circle calculation routines due to minor rounding errors
- Fixed problem with run-time errors in QtoH routine causing design evaluation to crash. Will now continue to analyze all designs, even if some cause run-time errors. Those causing errors get filtered out of final design evaluation report.
- Added this line 4/2/2002 to avoid error when running design alternatives on Courtland Canal flume: If BaseFlume.SillHeight <= 0 Then BaseFlume.SillHeight = 0.03
- 3/27/02 - Executable file stamp 1.05.0018
- Added discharge units of Megaliters/day and Megaliters/hour (Note that 1 Megaliter = 1 cubic dekameter). Cubic dekameters/hr was already an option in the program, but Australian users more frequently refer to this as Megaliters/hr.
- Added error trap in wall gage preview routine to catch the error that is raised when no printer driver has been installed.
- Modified new flume dialog to force units selection and entry of a flume description when creating a new flume based on the built-in default flume.
- 9/18/01 - Executable file stamp 1.05.0017
- Ensured that alpha1=1.04 everywhere (except when computing ideal discharge). A couple of places still used alpha1=1.0, which causes values of H1/L to be slightly different.
- Added validation of roughness height input to issue warning if user enters a value greater than 2 cm
- Added iteration counters and failed-convergence traps to detect convergence failures in HtoQ and QtoH routines caused by extremely large roughness values.
- Resolved some inconsistencies in behavior and error messages when user tries to open a FLUME 3.0 database file and instead selects a DBF file that is not a FLUME 3.0 database.
- Modified design module to assume that H1/L ratio will be within range 0.07 to 0.7 after design is refined, and estimate flow measurement accuracy accordingly when trying to find an acceptable design.
- Fixed bug causing subscript out of range error when creating back-to-back new default flumes.
- Fixed bug causing wall gage report to retain error messages from previous reports generated.
- Added check for changed flume when exporting data report and design report to clipboard.
- 2/21/01 - Executable file stamp 1.05.0016
- Fixed problem with display of very long path and file names in title bar of the application
- Improved design module so that it gives more helpful information when designs are limited by submergence. Previously it was not evaluating other design criteria once a submergence problem was identified.
- Improved several aspects related to the diverging transition of flumes that are solely width-contracted. Fixed several problems with screen display (bottom profile) for these kinds of flumes. Modified data report to give more tailored messages regarding expansion length of width-contracted flumes.
- Added flume description text to bottom profile. It can be edited there.
- Changed color of water surface profiles to a darker blue that shows up better in print.
- Added (bed slope) label next to "hydraulic gradient" on Manning's n tailwater calculation screen.
- Fixed minor problems with artifacts in the section drawings for U-shaped and V-shaped sections. Fixed similar problems with water surface profile displays (odd lines showing up in the tailwater section in some cases).
- 10/5/00 - Executable file stamp 1.04.0014
- Fixed bug in which WinFlume did not immediately recompute tailwater levels when the tailwater cross section shape was changed. The user had to re-open the flume/canal properties dialog to get the tailwater levels to recompute (assuming that one of the Manning's equation methods is being used).
- Added a reminder message box to prompt the user to change the flume description if it is still set to the default description, or if it begins with the words "Copy of".
- 10/2/00 - Executable file stamp 1.04.0013
- Rebuilt setup kit to address some DLL/OCX version compatibility issues.
- 6/13/00 - Executable file stamp 1.04.0012
- Fixed problem with calculation of area and top width in the trapezoid-in-U-shape section. This caused a discontinuity in the rating tables of flumes of the Palmer-Bowles type. This problem also existed in FLUME 3.0 and was never found until now.
- 4/17/00 - Executable file stamp 1.03.0011
- Fixed problem with display of X-axis on measured data graph when there are warnings in the rating table (overlayed graphs did not always have same axis ranges, so they didn't overlay properly).
- Fixed bug in linear interpolation tailwater data grid. Now discards any data in the file that would indicate two different tailwater levels for the same discharge.
- Added test for L/W ratio of throat section in flumes that are solely width-contracted. Error/warning number 24 will tell user that L/W > 2 is recommended for width-contracted flumes. This is based on preliminary lab research, and may change or be refined in the future.
- Rearranged controls used to enter data when calculating tailwater levels using Manning's n and the bed slope. Added descriptive label to help users pick up on the use of the expandable list of pre-programmed n-values.
- Changed all references to Q-H data for tailwater calculations to Q-y2 data for consistency with nomenclature being used in publications
- 3/8/00 - Version 1.02 - Executable file stamp 1.02.0010
- Fixed bug in section editing form. When setting a section equal to another section, the program was switching the display to the other section after completing the copy operation.
- 3/7/00 - Version 1.02 - Executable file stamp 1.02.0009
- Attempting to get a version with DLL's compatible with USDA-NRCS CCE computer systems. This distribution kit includes a newer COMDLG32.OCX control, and eliminates distribution of DBGRID??.OCX and GRDKRN??.DLL files in both the 16- and 32-bit versions of the program.
- 3/2/00 - Version 1.02 - Executable file stamp 1.02.0008
- Corrected error in calculations of derivatives for Newton's method in AStarCv routine. This problem was just introduced in the 1.02.0005 compilation.
- 2/26/00 - Version 1.02 - Executable file stamp 1.02.0005
- Returned to using Graphics Server (GSW) 5.0. GSW Version 5.5 was used in WinFlume 1.01, but was creating installation problems on some machines due to new DLL dependencies. GSW 5.5 did not make any material improvements in the program. Testing has shown that the WinFlume 1.02 executable files will work properly with either GSW 5.0 or 5.5, so users who have version 5.5 already properly working on their system should be able to run this version of WinFlume.
- Fixed problems in the AStarCv routine and StationaryDesign routine that sometimes caused design module to enter an endless loop. This problem was primarily exhibited with complex trapezoid control section. The AStarCv routine was comparing two estimates of A*/A1 and using Newton's method to get the estimates to converge on one another. However, Newton's method was not converging well because the relations are highly non-linear. I added some robustness to the convergence routine in AStarCv. On a related note, there is a typo in equation 6.52 in the 1993 FLUME book. It should be the formula for (A*/A1)^2, not just A*/A1. Problems with convergence of this routine only affected the design module in its attempts to find suitable designs. They did not affect the actual rating tables for designed flumes, since the Cv value is computed by a different method once the design is known.
- I improved the response of the FindMinMaxThroat routine when StationaryDesign is unable to converge on a valid flume design, which happens very rarely. If StationaryDesign fails, the program will now create the minimum and maximum-contraction flumes that can exist within the canal geometry, and then analyze everything between those limits.
- Fixed bug that prevented redraw of the section shape when changing from a trapezoidal to a rectangular shape while the base width of the trapezoid is set to zero (i.e., user has reduced trapezoid to a triangle. WinFlume will now set the width to 1 unit of length (dependent on units choices) when changing to a rectangular section if the base width of the trapezoid is presently zero.
- Added capability of using a fixed discharge interval for the rating tables used to develop curve-fit rating equations. This change required a change in the *.flm file format. This version of WinFlume will load all files created by prior versions of WinFlume, but older versions of WinFlume will not load the files created by version 1.02.
- Changed file format to make it possible for this version of WinFlume to load future flume files, even if future versions of WinFlume introduce new file formats. Essentially, the new file format and all future formats will be internally self-documenting, so that WinFlume 1.02 can load future file formats, whatever they may be, while ignoring any new data related to future enhancements.
- Changed brief design review report on main flume geometry screen so that it does not present the design criteria evaluation details if there were fatal errors at min or max discharge (they are usually meaningless since a head-discharge rating could not be computed due to the fatal error).
- Drop-down lists of default choices (e.g., head measurement devices and construction materials), will now expand by one item to include custom entries by the user. This makes it possible for the user to change to a different "default" device, then change back to their custom entry without having to manually re-enter the data.
- Rearranged data entry frames on first tab of wall gage form so that constant-head settings are on left and constant-discharge settings are on right. This corresponds spatially with the locations of the results shown on second and third tabs.
- Rearranged the freeboard requirements form so that the parameters and details of both options for specifying freeboard requirement are always visible. This makes it more apparent to the user what choices are available.
- Put a check in the equation fitting module to ensure that at least 6 data points are used for curve-fitting.
- Added a centimeter option for length units.
- Added buttons on the section-editing screens to make the section you are currently editing identical to one of the other cross-sections.
- Fixed a bug in the rating table graphs that occurred when trying to plot some 2 and 3-parameter plots with a range of the graph that has errors in it. Left axis of graph was sometimes garbled because overlayed graphs were not aligned with one another.
- Added tooltip descriptions to the text boxes used to enter bottom profile dimensions. Tooltips will open up when you click on a text box. To clear the tooltip, click on the text of the tip.
- 12/20/99 - Version 1.01 - Executable file stamp 1.01.0003
- Incorporated Graphics Server 5.5 service pack. Only visible change is that the Graphics Server icon now shows up in the System Tray, rather than in the taskbar.
- Rewrote the Pathname function to no longer use the Common Dialog control. Later versions of COMDLG32.OCX that may be on some users systems do not support the .Path property of this object.
- Version 1.01 (December 17, 1999) - Executable file stamp 1.01.0002
- Fixed problems related to the use of the comma as a decimal separator. Replaced the Val() function with the CDbl() function in most places because the Val() function is incompatible with the comma as the decimal separator. This is a documented bug in Visual Basic that I had not yet encountered. It manifested itself as a run-time error when attempting to generate the ditchrider's rating table, and also sometimes caused dimensions entered into the bottom profile and section-editing forms to be truncated to integer values.
- Improved navigation and editing of the measured data table and the Q-H lookup table for tailwater data. Tables can now be navigated using cursor arrow keys, or Shift-arrow keys for right-left movement. Pressing the down arrow when in the last row of the measured data table will automatically add a row to the table. Fixed some minor bugs in processing of data entered into those tables (new entries were sometimes being rejected by the program or converted to integers).
- Modified convergence test in normal depth routine to allow convergence in cases in which flow and channel size are extremely large. Old convergence test failed because the depth is calculated as a single-precision number, which makes it difficult to reach the pre-set convergence criteria when magnitude of flow is large.
- Fixed problem with rating tables and other items that use the grid control displaying backwards (fixed columns at right side of screen) on systems using a newer GRID32.OCX file, version 220.127.116.11. This required recompiling the program with the newer GRID32.OCX in place on my system, then updating the setup kit to distribute the new version of GRID32.OCX.
- Added ability to copy standard rating table and ditchrider's rating table to the clipboard in a tab-delimited format that is more easily pasted into spreadsheets. This would allow customization or further analysis of rating table data. Space-delimited format is also still an option.
- Added a slope indicator label on the converging transition in the bottom profile drawing.
- Disabled the Vary Side Contraction option on the design form when the control section is a V-shape.
- Made some minor improvements to the layout and appearance of the design review report.
Last reviewed: 05/11/12