TRNSYS 15 Fix List

Below are descriptions of recent fixes made to various pieces of the TRNSYS package in response to noted operational problems or inadequacies.

TRNSYS Patch 15.3.00-P01 (May 1, 2003)

Some information has been added to Type16 proformas to facilitate their proper configuration:
- The commonly used workaround to define the starting day (using an equation STARTDAY=INT(START/24)+1) has been described in the "more" button of Parameter 4 (Starting Day)
- Examples have been given to compute the Shift angle in the "More" button of parameter 7 (Shift). Users are now informed there that Type89 prints a notice in the listing file to help them configure Type16
- Examples have been added to the "More" button of Input 6 (azimuth angle) to help prevent the confusion between northern and southern hemispheres
Some information has been added to Type68 proforma (output variables description)
- More flexibility was added to Trnshell to control several groups with radio buttons. "//checked" is now added by TRNSHELL to the line of the radio button that is currently selected. If no "checked" instruction is found, TRNSHELL keeps using the former method (find out the selected radio button by looking at the commented and uncommented groups)

Bug Fixes
Type16 proformas:
- The azimuth angle description: "180=facing North" was only valid for northern hemisphere
- The description of horizontal radiation modes was wrong in "general description" and in the description of parameters (Par 1 "More" button)
- The units for diffuse radiation on tilted surfaces were wrong (W/m^2 instead of kJ/h/m^2)
- Type16a.tmf proforma had a wrong maximum number of surface (4 instead of 8)
- Type 68 was giving incorrect results when using two or more instances of the Type
Type68 proforma:
- The description of the azimuth angle was wrong (only valid for Northern hemisphere)
- The maximum values for the number of openings and the number of surface angles were wrong. TRNSHELL / TRNSED
4. Type89 proformas:
- Proformas for Type89 e,f,g,h (IWEC and EnergyPlus) had incorrect units for diffuse solar radiation and for extraterrestrial radiation (beam, diffuse and total)
- Type89 proformas b,d,e,f,g,h (IWEC and EnergyPlus) had incorrect dimensions/units for illuminance and luminance variables
5. DynamicData.for:
- A small problem was fixed in the labels for printing
- The program was not able to open normal deck files (.DCK) larger than 55KB with Windows 9x. .TRD files opened through the "TRNSED/Edit TRNSED File" menu were OK. TRNSHELL now uses the same editor for both file types.
- "Create executable" command: nothing was happening if no data was given for title and author. An error message is now displayed
- "Create executable" command: TRNSHELL did not append a ""\"" to the destination folder so wrong path/file names were sometimes used
- "Create executable" command: if the full name (including path) of the .trd file to be included exceeded the input field length, TRNSHELL was not including the file(s)
- TRNSHELL was sometimes erasing the .TRD file on exit when errors had occurred (e.g. missing picture)
6. Prebid
- The French layer library had some errors in the conductivity values (among others: carrelage, hourdis, Lame d'air, Laine de roche, laine de verre, acier)
7. IISiBat
- IISiBat was crashing when proformas with nonexistent dimensions/units were edited. It now changes the dimension to "any".
- IISiBat was not accepting case variations of Inf in the parameter range (e.g. INF or inf). This would cause IISiBat to reject parameter values even though they were acceptable.

Release 15.3.00 (March 18, 2003)

Warning: The Type56 and Prebid introduced in Release 15.2.07 require you to open your old building description files and to regenerate the BLD and TRN files before running them. Failure to do so will generate a FORTRAN error

- The online plotter now keeps more values when drawing the plot. This greatly improves plot zooming capabilities and makes the plots much smoother.
- Online plotter window now says "Exit online" and not "Exit" in the File menu.
- Right-clicking on the graph stops / resumes the simulation.
- Switching between different online plotters by clicking on the tabs does no longer stops the simulation.
- The Y axis scale is now controlled by a popup box that offers more flexibility than left click/right click method previously used.
- The RAM requirements of the online plotter have been dramatically reduced, allowing much greater precision in zooming.
- Users may now obtain variable values by holding down the SHIFT key while in zoom mode.
- The "Apply" button in the File/Setup menu has been renamed "Store" to more accurately describe its function.
- The "Create TRNSED Distributable" menu item has been moved from the File menu to the TRNSED menu.
- In TRNSED, the default behavior if no "TAB2" instruction is found now right-aligns the units of input fields, giving more flexibility in graphical design of TRNSED distributables.
- The standard Windows (TM) function used to save TRNSHELL/TRNSED files has been replaced with a custom designed function that better suits TRNSHELLs needs when flipping back and forth between "Source" and "TRNSED" views.
3. Source Code: Type2
- IISiBat proformas have been updated to better explain and stress the importance of the solver choice as it affects the controller mode.
4. Source Code: Type56
- The check for incorrect solar radiation incidence angle now prints a warning instead of an error message. The simulation goes on using a null beam radiation.
5. IISiBat
- The proforma implementation has been rewritten. It is now possible to use the Windows Copy-Paste functions in proformas.
6. Weather files
- New examples of weather files have been added. See \Trnsys15\Weather\weather_files.txt for more information

Bug Fixes
- Trnsys would exit immediately if run from the Developer Studio or from a command prompt without arguments when the last file opened was directly in a root folder (e.g. c:\project.dck, or d:\Project.dck)
- Y axes labels of the online plot were often displayed incorrectly.
- The time displayed by online plotter more accurately reflects the simulation time.
- Trnsys would sometimes crash or display empty plots due to a rounding error in calculating the number of points.
- THe PageUp button caused an Unhandled Exception when in TRNSED view.
- TRNSHELL was not clearing the memory it was using, eventually using a significant amount of RAM unnecessarily.
- TRNSED was showing a large gap in its display in some circumstances. For example, if a picture was deactivated by a radio button choice, the space necessary to display that picture was still shown.
- TRNSHELL was not saving the TMP file used for flipping between views properly when the simulation was run from "Source view".
- TRNSED would not run parametric tables. It was looking for a file instead of a
- TRNSHELL was inserting "Carriage Return" characters randomly in the TRD file.
- Text and picture alignment in TRNSED was not working properly. 
- Trnshell was not correctly saving changes when a user manually modified a radio button group in "SOURCE" view. 
- The "Find" function was not working properly in output and listing files 
3. Source Code: Kernal
- Trnsys was sometimes creating empty CNFGTR.TRN files that would then prevent it from running properly in subsequent simulations. All file "OPEN" instructions are now preceded by an "INQUIRE". Trnsys now uses default values and prints a warning if an empty or corrupted cnfgtr.trn file is found. 
- Long variable names (more than 10 characters) in equations could lead to unexpected results. An error is now generated when a user attempts to use long variable names
- The Type301 to 310 DLL's included in the UserLib folder were removed. They were not used and added to the confusion on how to use external DLLs and Type61. 
- A bug in Trace.for and Print.for was causing Fortran runtime errors in very long simulations (when the number of calls to a component exceeded 1.0e6. The maximum is now 1e9)
- A bug in sumary.for was causing Fortran runtime errors in very long simulations (more than 99999 hours). The maximum is now 9999999 hours
4. Source Code: Type2 
- In mode 0 ("new" control strategy), the use of more than 5 controllers was causing a range check error. 
- No check was made that the solver used corresponds with the control strategy (i.e. the value of NSTICK). Type 2 now issues an error and stops the simulation if the wrong solver is used.
5. Source Code: Type5
- The equation for cross flow configuration, both fluids unmixed had a typo.
6. Source Code: Type25
- The columns were not aligned properly when spaces are selected as column separator.
7. Source Code: Type28
- An error was generated for simulations longer than 99999 hours. The maximum is now 9999999.
Source Code: Type34
- The proforma did not include the fraction of solar shading output added with release 15.2.07. The output can be used as the "external shading factor" required by Type56
9. Source Code: Type55
- It was impossible to obtain a daily integration with reset time at midnight.
10. Source Code: Type56
- A bug was fixed for walls with active layers and identical boundary conditions
- A bug was fixed for Internal shading devices using schedules
- Latent loads (humidification/dehumidification) were dependent on the internal time step
- The relative humidity was not correctly initialized
- The transmittance was not correctly computed for internal windows with internal shading devices
- The humidity gains were not correctly initialized
11. Source Code: Type68
- Type68 was not checking that obstruction angles behind the object are at least as high as the skyline of the object itself.
- Type68 was causing a "range check error" if used with only one opening (fix in DynamicData.for).
- Type68 proforma contained an incorrect description of the input data file required by Type68. Two data lines were missing in the description and users following the directions would get run time errors. 
- The Type68 proforma description of unshaded and shaded fractions were reversed. 
12. Source Code: Type71 
- IAM outputs (Transversal and Longitudinal IAM's) were incorrect if several Type71s were used. Negative values could occur for the transversal incidence angle.
13. Source Code: Type72
- IAM outputs (Transversal and Longitudinal IAM's) were incorrect if several Type72s were used. Negative values could occur for the transversal incidence angle.
14. Source Code: Type73
- IAM outputs (Transversal and Longitudinal IAM's) were incorrect if several Type73s were used. Negative values could occur for the transversal incidence angle.
15. Source Code: Type80 
- Generated a Range Check error is more than 10 surfaces were defined.
16. Source Code: Type88 
- Type88 allowed the humidity ratio to go above saturation. In that case the energy balance was wrong. The humidity ratio is limited to to the saturation value and an output has been added with the energy from condensation. The proforma has been adapted to reflect the new output.
17. Source Code: Type89 
- Type89 was generating a Fortran File Access Error when the T89.TMP file was found to already exist.
- Type89 was not able to read EPW weather files with question marks in the "present weather observation" field.
- The shift angle printed to the listing file for IWEC files was calculated incorrectly. 
- Type89 was skipping one data line in multi-year simulations
18. Source Code: Type93 
- Output descriptions in the Proforma were confusing.
19. Source Code: Type96 
- source code has been modified to include outputs 10 to 13, which were referenced in the manual. Warning: the manual describes an "inverter losses" inputs that does not exist. Total losses given in output 10 are actually equal to the power dumped.
- The "external file" tab was asking for a logical unit number instead of a file name.
20. IISiBat
- Dictionary.dic (IISiBAt units dictionary) was missing in the official version
- IIsiBat.ini included a file on the u:\ drive (default deck name)
21. TRNSlate.exe
- The utility program to update Trnsys 14.2 deck files to Trnsys 15 had a problem with equations connected to inputs.
22. Examples
- Example 2 was using an incorrect deck file name
- Example 10 was using an incorrect version of the Type96 proforma
- sdhw.trd was using Type2 in the wrong mode with solver 0

Release 15.2.07 (July 17, 2002)

Type34: New output added (percent solar shading). It can be directly input to the type 56 as the window external shade factor 
2. Type55: he outputs of Type55 now keep their value until the start of the next period instead of resetting to 0 at midnight. 
3. Type56: an open source version has been added (\Types\Type56\opensource). This version is identical to the closed-source version except that active layers and internal windows are not available. Only one Version of TYPE56 (closed or open) can be included in Trnlib.dll
4. Type89 (standard weather files data reader) can now read ASHRAE IWEC files (mode 4) - undocumented feature!
5. TRNSHELL/TRNSED: some improvements have been made in "Windows application" behavior: mouse scrolling, Copy-Paste, Find, etc. The Windows "close" button (X) is now equivalent to File/exit and updates the recent files menu.
6. Prebid 5.0.8
 -  clarification of some points (role of the path to input file, etc.)
 - changed dimension of schedules time
 - Negative values are only allowed to be defined as an input in the dck. No constant negative value e.g. wallgain is allowed in the bui-file. This is now checked by PREBID
7. ExtDLL.for has been moved to the \UserLib folder for the sake of clarity. This file should not be added in the TRNLIB project, it is only used to create ExtDLL.dll and ExtDLL.lib
8. The included workspace for the Compaq Visual Fortran 6.6A (CVF66A) is now configured by default in Debug mode. The correct version of libraries are ignored (libc.lib or libcd.lib)
9. The online plotter displays more significant digits on the Y axes. This is espacially useful when small values as absolute humidity ratios are plotted

Bug Fixes
1. Various bugs have been corrected in TRNSHELL and TRNSED.
  - "I/O Error 104" is not caused anymore when using TRNSED in a distributable application. 
  - Radio button groups were not saved nor displayed correctly in some configurations
  - TRNSHELL was deleting the .TRD file in some cases
  - The error handling in TRNSHELL "Edit TRNSED file" mode has been improved and is more robust
  - Gaps were sometimes appearing in the TRNSED display
  - Parametric tables are functional
2. Prebid:
 - The wall layers order was reversed in the french wall library
 - There was a confusion between "heavy concrete" and "heavy wood" in some layers in the French wall library
3. Type47 proforma had an erroneous default value. Proformas for Type47b, c, d and e have been corrected
4. Type54 weather generator was causing file write/close errors when combined with Type56
Type55 periodic integrator was not working at all if the integration start/stop times were not simulation start/stop times. In particular the example in the manual was not working
6. Type56 bugs:
 - Static declaration of wbarint to 20 zones changed to allocated memory
 - Additional resistances of shading devices for single pane windows corrected, absorption of long wave rad(qradint) and absorbed rad on int. shading (qabsh) for single panes corrected.
 - Edge correction effect on heating load corrected
 - Active layer added coupling for regular zone FCLPG=4.12 which was wrong
An external file tab has been added to Type70 proforma
8. Type80 proformas had various errors
9. Some units were incorrect in Type89b.tmf and Type89d.tmf
10. Type90 various bugs: 
  - Non-zero power output was sometimes given for zero wind speed
  - TRNSYS S storage array was not allocated correctly, which was causing errors in simulations, sometimes affecting other Types using the S array
  - Ambient temperature was treated as 10th of °C and not °C
11. Type95b and Type95c proformas were erroneously asking for an input file. This was causing a Trnsys error even if no file was selected (tries to assign unit 1)
12. Example1 has been corrected (error in link between Type16 and Type1)
13. Spread.exe has been added to the Presim distribution
14. The number of constants used in equations has been increased and now changes when the maximum number of equations is increased
15. Convergence error messages have been modified to cope with Type numbers larger than 99
16. Equations: when a negative number was raised to a non-integer power, the value was left unchanged and a warning message was printed. Now, an error is generated and the error message points out the concerned equation
17. DynamicData.for: When the 4th dimension of X values was added, the way of calculating the size of the YDATAdd( ) and DATAINdd( ) arrays was not modified
18. Type93 proforma had only one parameter. A second parameter (Nb of inputs) has been added

1. Begin.dck, Fixed.dck and Example1 have been corrected (error in link between Type16 and Type1)
2. The restaurant example was including some walls that were confusing regarding the layers order. These walls have been replaced.
3. IISiBat Example 5 (Ex5.tpf) had blank initial input values for Type28 that were causing range check errors when "write TRNSED commands" was set to "true"

1. The wise installation wizard was telling to add the include folder, while this folder is now included in the updates to prevent compatibility problems.
2. The Readme file is now available in TXT and HTML formats and includes this fixlist


Release 15.2.02 (January 24, 2002)

1. Trnsys now uses default values for the parameters in cnfg.trn if the file is not found instead of generating an error.
2. The \Trnsys15\Workspace\CVF66A folder includes a new workspace adapted to Compaq visual Fortran 6.6A. This workspace implements the new compiler settings and includes three projects (Dummy library, External DLL call and Trnlib DLL).
3. TRNSED has a new feature. Opening a file with the "Edit Trnsed file" command gives acces to a dual view (source code / Trnsed view)

Bug Fixes
1. New compiler settings have been implemented in order to improve the compatibility between Trnsys and some localized Windows versions.
2. 3 bugs have been corrected in Online.for: first time step was printed twice, typeck was called at each time step and the new label size (16 characters) was not correctly handled
3. Several errors in the Type71 proforma have been corrected, as well as in Type16g and Type16h proformas.
4. Some errors in the German prebid library (prglay.lib) have been corrected.
5. A bug has been corrected in Type19.for: It was not possible to use the maximum number of surfaces (15)


Release 15.2.00 (December 6, 2001)

1. it is now possible to have a dynamic deck name meaning that if you rename one of these decks, then the files that you have selected, will be renamed as well. To make use of the feature, put three * characters before any file extension that you wish to be dynamically named. For example: ASSIGN ***.lst, ASSIGN ***.out. TRNSYS will append the path and deckname onto all files that you have ASSIGNed in this manner.
2. You can pause a simulation with an online plot at a specified time. To make use of this feature, run the simulation and pause it (F7). Next, go to the calculation menu, and select the item “Pause at…” You can now enter a simulation time and unpause the simulation (F8) and it will stop at the specified time. The information is stored in the .pti file for the deck so is remembered from one simulation of the same deck to the next.
3. Type65 online plot labels are now limited to 16 characters instead of only 8. As part of this enhancement, the TYP65 common block has been removed from TRNSYS.
4. TRNSYS now remembers the most recent browse directory.
5. DynamicData has been expanded to include 4 dimensions. DynamicData is called in the same manner as Data and will eventually replace data entirely.
6. TRNSYS Type numbers 1 through 999 are now called (previously only Types 1 through 300 were called)
7. Prebid: user id files with accented characters in them are now accepted.
8. Users are no longer obliged to send out their own files with a TRNSED application. A new Make Distributable menu item has been added to TRNSHELL which creates a file for specific TRNSED applications. The feature may be complicated to use so it is initially disabled and users are asked to contact their distributor to activate it.
9. The "Make TRNSED" command now copies all required files into the destination folder

Bug Fixes
1. Unit numbers of the form x0y were written to the list file as xy – a new proc.for fixes the problem.
2. Type36.for - The coefficients C1 and C2 were different from the typical values suggested in Solar Engineering of Thermal Processes.
3. Online – the label colors in the zoomed online did not match the trace colors on the right axis.
4. psych.for – the new scheme of forcing an iterative calculation of wet bulb temperature at every timestep revealed some insufficiencies in the iterative scheme. An exit was added to the wet bulb iteration if the slope of the error function gets below 0.0001.
5. The '/' characters in the pathname to units.lab in rcheck.for are forcing the file to be in the same directory as the deck.
6. The array bounds in state.for were too low and have been increased.
7. Type34.for - Type34 could output negative radiations in certain cases. It now checks for negative radiations and zeros them.
8. Various IISiBat Proformas had typographical errors, unreasonable initial values and some unit errors. The affected proformas are: Type55.tmf, Type93.tmf, Type16x.tfm (all type16 proformas), Type80x.tmf, Type28b.tmf, Type88.tmf, Type60x.tmf.
9. Examples affected by the modified IISiBat proformas have been updated.
10. Type89.for – Type89 was not reading enough characters per line in EnergyPlus weather files.
11. Prebid now correctly handles the material names in the American library without causing materials to be written twice in the .bui file.
12.  In Prebid, multiple internal windows with the same area in an adjacent wall caused an error.
13. In type56, NTYPEs 51 and 56 g-value outputs were in W/m2 instead of kJ/hr-m2.
14. An error occurred with the internal timestep in humidity coupling calculations.
15. TRNSED was checking for the existence of all ASSIGNed files before running, erroring because it did not find the .out files. It now gives a warning message.
16. It was not possible to type in a file name in TRNSED, you could only modify it using the browse box.
17. TRNSED now runs correctly if its required files have spaces in their filenames and pathnames.
18. Some Iisibat proformas included a variable name starting with #, which lead to errors when generating a Trnsed deck. Proformas for Type8, Type8a, Type8b, Type11a, Type11b, Type20, Type57, Type65 and Type65b have been modified. Iisibat examples using these types have been updated.
19. TRNSYS.exe no longer accepts TRNSYS 14.2 style files.
20. Using the same UNIT number twice caused a Fortran error that was poorly handled.
21. The C1 and C2 coefficients in Type36 were different than those in mentioned in Solar Engineering of Thermal Processes.

1. There are minor errors in the main manual table of contents. Section 4.3.4 is listed twice. Section 4.9.7 missing.

1. The shortcuts created for Iisibat are now working if installed on another drive than C:\


Previous TRNSYS Releases

June 27, 2002 - Enhancements: 1. new Type68 which reads a SimCad (or hand) generated file of architectural mask data and provides Type56 with shading values for affected openings 2. a new dynamicdata.for exists which will grow to accomodate whatever size data file users wish to read. 3. New Type66 which is significantly faster than the previous version. The new version uses the Windows clipboard for communication instead of reading and writing files. Bug Fixes: 1. Corrected building.tpf and Example2x.tpf files. 2. Type95.for - error in setting the MODE if the year was 2001. 3. Type71.tmf - corrected an error in the proforma. 4. Corrected errors in the French wall libraries. 5. Translated French window libraries. 6. Removed broken help link in TRNSHELL. 7. Rcheck.for - increased bounds in OCK and YCK arrays to prevent a floating point error 8. Readin.for - increased array bounds on ARRAY arrays to prevent a floating point error. 9. Type34.for - fixed an error in setting outputs that was causing negative radiation values. Documentation: 1. screen capture in I3 manual said TRNSYS 14.2 (p. 78) 2. Incorrect Prebid directories were referenced in the I3 manual (pp. 73 and 74) 3. Added a description of Type25 optional PAR(6) which controls tabbed or spaced output. (main manual p. 4.10.1-2) 4. corrected error in Type25 main manual which said that only 10 inputs are possible (p. 4.10.1-1) 5. New documentation for the updated EES connection (main manual document 4_1_9.doc) 6. Documentation for new Type68 (main manual document 4_9_8.doc) 7. Because of the additions, it is not possible to just reprint some pages of the Volume II Component Descriptions manual so you will need to reprint it entirely.

April 5, 2002 - Enhancements: 1. default absolute and relative tolerances changed to 0.001 0.001 2. default limits for iterations per warning and warnings per error changed to 30 30 3. added 4 new modes to the Type5 heat exchanger for 1. crossflow with hot side mixed 2.crossflow with both fluids mixed 3. crossflow with neither fluid mixed 4. shell and tube 4. all compiler dependencies are removed from trnsys.for to trnsys_win32_cvf61.for. Other versions of the file (for other compilers and operating systems are in the \oscomp directory. 5. TRNSHELL now allows for different fonts and font sizes 6. Psych.for has been updated to ASHRAE 1997 standards. 7. Enhancements to the Type80 heat transfer corellations Bug Fixes 1. data.for - a 2D data file with only one value in the 2nd dimension was not read correctly 2. begin.dck - didn't create output files because the printing time and simulation times didn't match 3. Type80.for - the INFO array declaration was not compatible with the Lahey compiler 4. Type96.for - an array was not subscripted correctly 5. Type72.for - 1. problem in passing YHOLD( ) to YCHECK( ). 2. problem in the call to DATA for reading Fr(ta) data. 6. proc.for - TRNSYS now checks the validity of the START and STOP times (before, you could put in a stop time less than the start time and cause a nasty Fortran error) 7. Type87.for - the PAR array was dimensioned to 3 and then filled with 6 values. This should have been a runtime error but it caused a compiler crash in Windows98. Patch A for CVF6.5 fixes the problem. Documentation Changes 1. a readme file has been added to the weather directory explaining the various files and their formats. 2. Introduction to TRNSYS with IISiBat: line 11 page 46 mentions a Type9 which should be Type89 3. Introduction to TRNSYS with IISiBat: line 6 page 43 mentions a Type9 which should be Type89 4. Introduction to TRNSYS with IISiBat: p 29. 1: data reader is classified as a utility component, not a physical phenomena component. 5. Main Manual: DATA file format example on p. 4.2.3-5 was wrong 6. Volume II Component Descriptions: DATA file format example on p. 179 was wrong 7. Main Manual: Type94 PAR(8) units are listed as A/K and should be V/K 8. Main Manual: Type94 output description list has units of kJ/hr listed - should be W 9. Main Manual: p 4.1.1-5 examples don't have headerskip PARAMETER in them. 10. Reprint page 19 (Type5 Heat Exchanger) of vol2comp.doc 11. reprint page 81 (Type27 Histogram Plotter) of vol2comp.doc (new footnote regarding LABELS for Type27) 12. reprint document 4_7_1.doc (Type 5 Heat Exchanger) of main manual 13. reprint page 2 of 4_10_3.doc (new footnote regarding LABELS for Type27)

January 16, 2001 Bug fixes: 1. Type73.for: Both TIN and TI were used for inlet temperature variables. 2. fixed the problem that Marion originally brought to my attention which was that big simulations running parametric tables over midnight would crash. 3. Type54: If a user had a data file with the wrong number of data items in it, an error was generated but the Type was called once again after the error had been written, which was causing a fortran error as opposed to exiting cleanly. 4. Type94: Providing ambient temperature in C not K to the photovoltaic array caused a TRNSYS crash instead of generating an error. TRNSYS error 203 is now generated if the input temperature is below 50 5. Type54: The radiation values from Type54 were shifted by 1 timestep because on the first timestep, the LAST array (which feeds OUT) was set to all 0s. This led to some high radiation values on west facing surfaces. The proposed workaround for users who do not wish to download the fixed version yet is to add a SHIFT value to the radiation processor equal to one timestep. Thus if a user was running an hour timestep, the SHIFT value would be 15 degrees. 6. Type94: There was confusion between the documentation and the code as to PARAMETERs 18 and 19. The code was fixed to reflect what the documentation said. 7. Type94: INFO(10) was set wrong. This had no effect on simulations unless the user was doing a TRACE on Type94. 8. Type94: Entering a large negative value of dI/dV@Isc resulted in a math error. TRNSYS now generates error 204 to warn the user if he/she enters a value of Isc and dI/dV at Isc which would result in a Voc of less than the one specified - which is physically impossible. For a complete explanation, look at the error.hlp file under error 204 - it has a picture. Enhancements: 1. modified the online plot trace colors so that there is no black or white line (one of which was always invisible depending on the background color). 2. subdivided the weather file directory into directories for different file types. Modified examples accordingly 3. added a new COMMON block to the trnsys source which carries the deck name. 4. You can use a 3 digit LIMITS value in trnsys now - there was previously a problem with writing to the LIST file (wrong data type) Documentation Errors: 1. main manual p 4.1.1-5 examples didn't have the new headerskip PARAMETER in them. 2. main manual p 4.11.5-14 Type94 output description list has OUTPUT 3 units of kJ/hr listed - should be W 3. main manual p 4.11.5-13 Type94 PARAMETER 8 has units A/K listed - should be V/K 4. component manual p 179 data file format was wrong 5. main manual p 4.2.3-5 through 4.2.3-7 same problem as above.

December 1, 2000 - Bug Fixes: 1. Type 9: If a data file had to be rewound during a simulation and the user was doing a formatted read (last PARAMETER > 0) and there were header lines in the data file to be skipped, then the simulation crashed after rewinding the data because the headerlines weren't being skipped the second time through.2. Type 9: If Type 9 was set to mode 1 (first line of data corresponds to simulation start time) and was a formatted read (last PARAMETER > 0) and the data file had header lines, then the sim gave a file reading error because type9 wasn't skipping the header lines. Enhancements: 1. New feature in Type 28 In simulaitons lasting longer than 1 year, Type 28 was not designed to provide sums after each year, only at the end of the simulation. You don’t want to write a sum after every December because you might be running a one year simulation from July 1 to June 30. Since, however long you run a simulation in which results are printed monthly, the Type28 results are grouped by 12s, I added a statement that provides SUMS after every group. 2. TRNSYS 15 help files. Typeindx.hlp, trnsed.hlp, trnshell.hlp and error.hlp.

November 3, 2000 - 1. New Prebid.exe. 2. New prglay.lib files for American and French libraries. Material names cannot contain certain characters and all of the libraries contained them. We've gotten rid of the illegal character names now. 3. New "copy" feature in the TRNSYS online plotter which will make the printing of online plots easier.

October 13, 2000 - 1. A bug fix in Type30: collector slopes of 90 degrees don't kill the simulation anymore. 2. A bug fix in Type34: collector azimuths greater than 360 degrees don't kill the simulation anymore. 3. A bug fix in fluids.for: unit conversion errors. 4. A fix to the building example. 5. Updated ModelConfFile.conf for iisibat3. 6. New IISiBat3.exe 7. New Prebid.exe with the following fixes a. COUPLING airflow was limited to 0 - 100 kg/h => now 0 - 100000 b. Warning "40 surfaces per zone not allowed ... " skipped. c. bugfix with "used types only" option deleted comfort inputs. d. active layer in ajacent walls with back first (big bug) 8. New Type56.obj files. 9. Removed w4-libd.lib file from the \trnsys15\prebid\lib\German directory. Updated all w4-lib.dat files. 10. Added a cleaner looking trnslate_instructions.txt file. 11. Removed link.fp4 from the distribution folder. You can recompile TRNSYS using Powerstation if you are willing to give up Type56, Type61, Type66 and the callprogram feature. All these routines use Fortran features that are not supported by FPS4.0.

September 8, 2000 - 1. A new TRNSYS.EXE. If you told the online plotter to run a 1 year simulation and that you wanted 52 plots (one per week), the online would look like it overran the end of the simulation because 8760 hours does not divide evenly into 52. With the fix, if the number of plots requested does not divide evenly into the simulation length, then TRNSYS modifies the number of plots to accomodate the simulation. 2. Big numbers were getting written to the automatic online plot output file as "********". I've fixed it so that they are correctly written. 3. A DATA statement typo in system.for was fixed. 4. Removed TABS from refprop.for, Type61.for and prep.for. 5. A bug fix in Type 16. The problem arose when radiation smoothing was curve fitting and the current time, minus the timestep was equal to sundown, causing a divide by zero. 6. A bugfix to correct problems using TIME as an INPUT to Type42.

August 17, 2000 - 1. IISiBat3 proformas- A correction was made to the Type93 proforma. Missig proformas for type89 mode 3 were added. 2. New TRNSYS.EXE which contains a "copy" command in the online plotter window for easier export of online graphics. 3. New DFIT.FOR and DINVRT.FOR kernal subroutines. These were needed because previously, the DINVRT subroutine was part of DFIT. 4. A bug fixed in Type94.for (photovoltaic module). Given a certain combination of PARAMETERS, the algorithm that searches for a the current and voltage at maximum power went out of bounds and tried to take the log of a negative number. 5. A bug fixed in Type60.for. If you have multiple type60s in a single simulation, you should definitely download this fix or the heater enable signals may get confused between the two tanks. 6. Removed all TABS from the source code making recompilation with the Salford Fortran for Linux compiler easier.

August 9, 2000 - 1. Made minor modifications to the source code in order to allow for recompilation using Fortran PowerStation 4.0

July 31, 2000 - 1. New clock.dos file - used in recompiling TRNSYS as a platform independent executable. 2. New Type56.obj files that further repair the humidity coupling problem (see July 6 posting) 3. Updated iisibat3 directory.

July 26, 2000 - 1. New steam.for which fixes a runtime "array out of bounds" error. 2. Updated trnsys.dos file (used to recompile TRNSYS as an exe. 3. New TRNSHELL which loads the new dck when a "Save As" command is executed.

July 6, 2000 - 1. New Type56.obj files which fix the problem that coupling humidity calculations were only correct for the last zone. For the other zones the humidity gain due to couplings was set to zero.

May 5, 2000 - 1. Source code fix in TALF. TALF can now accept a cover material index of refraction of 1. without getting stuck in an endless loop. 2. Source code fix in Type36. Type36 no longer errors when the outermost wall node temperature and the ambient temperature are the same in Mode 4. 3. Revised Type80.tmf files for iisibat 3.

May 4, 2000 - 1. Updated Type80.tmf files for iisibat 3.

NOTE: Changes made prior to May 1, 2000 are not documented here. Please send email to if you have questions about these changes.