Troubleshooting - Toolbar Error List

In This Topic

Error Analysis

General

After you have done a calculation, you will be interested to know whether it was successful or not. There are three sources that tell you about the last calculation:

the message box displayed after the calculation
the "Error List" toolbar
the calculation information sheet

Once a calculation has been completed, Ebsilon displays by default whether the calculation was successful, how long it took and how many errors or warnings there are.

It contains detailed information about the success of your calculation. It is possible that

the calculation could not be done at all e.g. because the input data were missing. In this case, the error analysis tells you what is missing and you have to specify the missing values,
the calculation could be started, but did not converge within the specified limit of iteration steps. In this case, take a look at the convergence diagram to check whether it is reasonable to increase the limit or eliminate the origin of the convergence problems, e.g. controllers that work against each other. The error analysis can be used to figure out the regions with the lowest convergence.
If the convergence target is not reached, the line with the highest deviation is highest deviation is highlighted in an error or warning message and the amount of the deviation is shown.
the calculation was done, but certain errors appeared
the calculation was successful. In this case, there may be warnings or comments.

If the message box has closed, you can also look up this information in the calculation information sheet.

Under “General Options“ à “Calculation“ it is possible to set if this message is to occur

always
never
only in the case of errors
only in the case of errors and warnings

The Error List toolbar opens automatically, if the calculation fails.

You can also open it manually e. g. with the menu command "CalculationàError Analysis". It contains all the messages of the calculation kernel: errors, warnings and comments.

Calculation information sheet

You can open the calculation information sheet from the menu command "ExtrasàModel Options", page "Results". it contains summary information, such as the type and the state of the calculation, number of iteration steps and the run time.

For more information, see "Model Options-Results"

The Error List toolbar

This toolbar can be opened e. g. by the menu command "CalculationàError Analysis". If a calculation ends with at least one message of the class "error", this window pops up automatically.

All errors are displayed in tabular form.

When it is displayed, the colour of the model turns from coloured to gray.

In the top line, the number of

"Errors",
"Warnings", and
"Comments"

is displayed.

By clicking, the messages of the corresponding category are shown in the table.

While you are inspecting messages, you can edit your drawing as usual.

For each class, there are certain types.

Errors can be

P/T/H definition errors
Missing start value
Missing mass flow
Missing pressure
Missing mass flow
Negative mass flow
Negative pressure
Negative enthalpy
Component error
Over-determination in mass flow (error)
Over-determination in pressure (error)
Over-determination in enthalpy (error)

Warnings can be

Over-determination in mass flow (warning)
Over-determination in pressure (warning)
Over-determination in enthalpy (warning)
Component warnings

Comments are

Component comments

You can read about the meaning of the error classes in the description of the simulation.

Note that over-determination can be an error as well as a warning:

It is an error if there is a contradiction between two or more redundant equations in the matrix solution. In this case, one of these equations is discarded, and the result is arbitrary, depending on the order of equations. This cannot result in a correct solution.
It is a warning if there are redundant equations in the matrix solution, but the redundant equations are equivalent. The calculation is correct in this case, but an error may arise if some input values are changed, e.g. in a part load case, and the redundant equations are no longer equivalent.

Intentional over-determinations that arise due to redundant measured values in case of a validation are neither an error nor a warning and will not be reported here.

In the case of over-determination, all components that set the overdetermined variable are highlighted.

For each type, there are certain objects that are affected by this type of error. In the list of error messages, there is one entry at the beginning of the corresponding type that displays all objects with this error. When this message is displayed, all objects are coloured that are affected by that type of error.

For component reports on balance sheet violations, both the absolute and the relative size of the violation are indicated.

The fourth button “Pipe Residuals“ allows to display lines for which there were changes between the second last and the last iteration step

that transgressed a definable threshold. This button serves as a combo box for selecting whether residuals in the mass flows, in the

pressures, or in the enthalpies are to be displayed, or also if all residuals are to be displayed. The convergence limit set in the model can

be used as a threshold, or the limits for each quantity can be entered by selecting “Edit thresholds”.

Moreover there are buttons in the top line that allow to

reset the display of the model to total view
activate the convergence diagram
activate and deactivate the error display (gray representation of all components and lines not affected by the error)
close the dialog
activate and deactivate an automatic closing of the dialog when carrying out a calculation

When clicking on the column headlines in the following line, the table is sorted according the respective size, first in ascending order, and

when clicking again in descending order.

By selecting one or more lines of the table, the corresponding objects in the model are displayed in colour.

In the error list, equation list, and measured value list you can select the corresponding component in the model.

The right mouse key allows defining which action is to be triggered by double-clicking:

Opening the properties window of the respective component
Selecting the respective component
Enlarging the respective component
Selecting and enlarging the respective component

Treatment of Topology Errors

The check of the topology for the connection of logic lines, set point value lines, and actual value lines has been tightened. This may result in error messages in the case of some models that previously calculated without an error message.

The following problem cases are detected:

Ambiguous structures (when there is more than one possibility in which direction a value is passed on)
Ring-shaped structures (this will also result when an actual value line and a logic line are connected at their ends as in this case it is not clear whether the actual value line is supposed to take over the value from the logic line or the other way round)

In most cases, these errors will have been caused by unintentionally snapping a free end of a line with activated snap radius. Users who often work with logic constructions are therefore recommended to deactivate the snap radius (under General Options --> Edit).

Strategy for errors that are difficult to localise

During the creation of a model, situations often arise in which the calculation (simulation) displays one or more errors.

The toolbar "Error List" is used to localise the errors.

In most cases, this is sufficient to eliminate the errors. Sometimes, however, a detailed determination is required.

A frequently used strategy is

cutting open streams and specifying individual variables (pressure, enthalpy, mass flow) with a signal transmitter (component 36), or also
dividing the entire model into sub-models.

This task is supported by the following two commands:

A) A stream can be ‘sliced’ if it is selected and its context menu ‘Cut open Stream’ is selected via the right mouse button. The stream is cut into two parts at the position of the cursor. The stream values are copied to both parts.

B) For the

general input value (component 33)
the boundary input value (component 1)
the measured value input (component 46)
the automatic connector (component 132) and
the limiter (component 147)

their context menu entry ‘Take over stream results to selected component’ is available for the general value input.

The pressure, enthalpy and mass flow of the stream are copied to the default values of the component.

For component 147, the default values INIM, INIP and INIH are used.

With these tools, it is possible to cut out parts of the model and determine them separately.

Evaluation of the convergence

The button "Convergence" can be used to call the convergency diagram directly from the error dialog. This is equivalent to the usage of "Extras"--->"Diagrams"-->"Convergence Diagrams...".

Additionally, the error analysis can be used to evaluate the degree of convergence in the streams of the cycle. To do this evaluation, click on the "Stream residuals" button in the error analysis window. An input dialog appears where you can specify limits for the precision of mass flow, pressure and enthalpy.

In this example, it is specified that all pipes shall be displayed that have a relative deviation of 0.00001 in the mass flow or in the pressure or in the enthalpy.

When you click on "OK", warning messages appear for each pipe where the relative deviation for the considered quantity between the two last iteration steps exceeds this limit. For these message, three additional error types are provided:

Lower precision in mass flow
Lower precision in pressure
Lower precision in enthalpy

You can view these messages, like all other messages, by going through the warnings. The message includes the actual value of the deviation for the corresponding quantity.

This is helpful in figuring out the source of a convergence problem. You can use this in addition to the convergence diagrams.

Note: If you want to figure out the convergence situation of a specific iteration step, set the limit of iteration steps temporary to that number and do a calculation. The calculation will stop at this iteration step, and you can evaluate it.

Impressum Disclaimer