Must be an odd number so that there is always a mid streamline.

Typically n = 9 or 17. If n_sl = 0, then 9 will be used.

Maximum n_sl = max_n_sl = 29.

If the mixing model is being used, (i_mix > 0 in section 4) then there has to be a minimum of 9 streamlines.

Typically specified as 500 but fewer are generally needed for simple radial compressor calculations to attain convergence.

If max_it_main < 4 then max_it_main = 4, so that at least 4 iterations are always done as a minimum.

If max_it_main = 0 then max_it_main = 500.

If the flow reaches the convergence limit before the maximum number of iterations is reached then the calculation is automatically stopped earlier.

Typically 10 and if max_it_mass = 0, then max_it_mass = 10.

If max_it_mass < 5 then max_it_mass = 5; experience shows that this is a sensible value.

If max_it_mass > 20 then max_it_mass = 20.

If the mass flow convergence tolerance at a particular quasi-orthogonal is reached before the maximum number of iterations are completed then mass flow iteration is stopped early.

If set to 0 or 1, then a single point is calculated.

Maximum number of points is 11.

If set to a value greater than 11, then a value of 11 will be used.

If set to 0 or 1, then a single speed line is calculated.

Maximum number of speed lines is 11.

If set to a value greater than 11, then a value of 11 will be used.

If i_progress = 0 then no intermediate information is printed.

If i_progress = 1 then intermediate progress of the iterations are printed to the history file.

If i_progress = 2 then data is printed to the history file and to the screen.

If i_display = 1, then no plot files are produced.

If i_display = 0 then the output files of type .txt are produced for display of the results with Tecplot.

If i_display = 2, then a comma separated variable file with extension .csv is produced for display of the results with CFD-Post.

If i_display = 3, then both i_display = 0 and 2 above are activated.

Other output formats can be incorporated as requested.

If i_restart = 0, the restart file (prefix.rst) will be used automatically if it is present (a warm start) and its content will be overwritten automatically at the end of a normal calculation. Note that this also overwrites the restart file even if the iterations are not converged, so that a second start with the same number of iterations starts with a better approximation. This is the normal way to use the program. The restart file includes the number of quasi-orthogonals and streamlines. If i_restart = 0 and this number has changed then the program makes a cold start with its own estimate of initial conditions (as in i_restart = 1).

If i_restart = 1 the restart file (prefix.rst) will not be used even if it is available and the program will set up its own initial conditions (a cold start). The content of the restart file will be overwritten as under 1 above. This is not generally recommended but can be useful in debugging difficult cases. This is equivalent to deleting the existing restart file and using option 1 above.

i_restart = 2 and 3 are special options for running the program when coupled to an automatic optimizer. In these cases the use of a good restart file reduces the number of iterations needed and brings a reduction in calculating time. Unfortunately some of the geometries being examined may be poor and so it is inadvisable to overwrite the restart file with poor results. If i_restart = 2 then the file runs with a restart file called best_restart.rst and writes the results on to prefix.rst. If i_restart = 3 then the file runs with a restart file called best_restart.rst and also writes the results on to the same file best_restart.rst. In both cases if the restart file best_restart.rst is not present then the internal initial estimate is used (a cold start).

If i_interface = 0 then no output interface file is generated.

If i_interface = 1 then the prefix.int file contains a summary of the results for use in radial compressor optimization.

If i_interface = 2 then the .int file contains the coordinates of the streamlines of the calculation.

If i_interface = 3 then the .int file contains the coordinates of the streamlines of the calculation in a format suitable for the stream file of the blade-to-blade program MISES.

The last digit controls the choke calculation mode. Using a value of zero for the last digit causes the choke mass flow limitation to be eliminated which may be more robust in difficult cases. Calculations of new cases should start in this mode.

Using a value of 1 for the last digit, enables the choke mass flow limitation calculation. This requires exact data for the throat areas to be specified and should only be used if this is available. It is necessary to set this value to 1 when using iteration to pressure ratio in choked stages.

The second-last digit controls the blending function calculation for the deviation between the blade angle and the flow angle as follows:

0 - Turbines use departure angle at the leading and trailing edge ends, Compressors use swirl at the leading edge and departure angle at the trailing edge.

1 - Turbines and compressors use the departure angle for the leading and trailing edges.

2 - Turbines and compressors use departure angle at the trailing edge and relative swirl at the leading edge.

3 - Turbines and compressors use departure angle at the trailing edge and the absolute swirl at the inlet.

4 - Compressors use swirl at the leading edge and departure angle at the trailing edge; turbines use swirl at the outlet and departure angle at the leading edge.

5 - Compressors use swirl at the leading edge, departure angle at the trailing edge; turbines use swirl at the trailing edge and departure angle at the leading edge.

i_flow = 0 to 4 is for a calculation with a specified mass flow. i_flow = 5 to 9 is for a calculation with a specified pressure ratio. See below.

i_flow = 0 then ref_mach, ref_phi, and ref_d, are specified, but if ref_mach > 3 then it is interpreted as ref_u, so this is equivalent to ref_u, ref_phi, and ref_d.

i_flow = 1 then ref_n, ref_mass, and ref_d are specified.

Note: If there is more than one spool in the calculation with different rotational speeds, then this is taken into account as follows:

- If i_spool = 2 and i_flow = 1 then ref_n1, ref_n2, ref_mass, and ref_d are specified.

- If i_spool = 3 and i_flow = 1 then ref_n1, ref_n2, ref_n3, ref_mass, and ref_d are specified.

i_flow = 2 then ref_n, ref_volume, and ref_d are specified.

i_flow = 3 then ref_u, ref_mass, and ref_d are specified.

i_flow = 4 then ref_u, ref_volume, and ref_d are specified

i_flow = 5 then ref_n, ref_mass, and ref_d are specified together with the ref_pr (total to static pressure ratio between inlet plane and the last trailing edge on the mid-streamline). The value of ref_mass is a start value for mass flow in the iteration to pressure ratio and has no effect on the final solution.

i_flow = 6 then ref_n, ref_mass, and ref_d are specified together with ref_pr (total to static pressure ratio) together with n_p_te (the total number of trailing edges at which a guessed value of the static pressure ratio is specified, followed by the guessed pressure ratios at each trailing edge, including the last, which is also defined by ref_pr.

i_flow = 7 then ref_n, ref_mass, and ref_d are specified together with the ref_p (static pressure at the last trailing edge on the mid-streamline). The value of ref_mass is a start value for mass flow in the iteration to outlet pressure and has no effect on the final solution. This option may be useful for low speed devices where pressure ratio becomes indeterminate.

i_flow = 8 then ref_n, ref_mass, and ref_d are specified together with the ref_p (static pressure at trailing edge pane) together with n_p_te (the total number of trailing edges at which a guessed value of the static pressure is specified, followed by the guessed pressures at each trailing edge, including the last, which is also defined by ref_p.

i_flow = 9 then ref_n, ref_mass, and ref_d are specified together with the ref_cu (absolute swirl velocity at the last trailing edge on the mid-streamline). The value of ref_mass is a start value for mass flow in the iteration to outlet swirl and has no effect on the final solution. This option may be useful for turbine calculations where the last blade row is a turbine rotor.

The versions with specified static pressure and specified outlet swirl have not been fully tested.

Other options are available for debugging purposes but are not described here.

The geometry definition of Vista TF assumes clockwise rotation. This leads to a certain convention for the sign of the blade angles (see Appendix A: Sign Convention for Angles and Velocities in Vista TF). In some cases you may have a counterclockwise machine with blade angles of the opposite sign. To avoid the need to change all the angles specified in the .geo file, an option is provided whereby the value of i_flow is given a negative sign. As a more sophisticated alternative the blade speed may be defined as negative.

If i_fluid = 0 then ideal gas with constant specific heats.

If i_fluid = 1 then liquid.

If i_fluid = 2 then a real gas calculated with the Aungier-Redlich-Kwong equations.

If i_fluid = 3 then real gas calculated with the Redlich- Kwong equations.

If i_fluid = 4 then an ideal gas calculation with variable specific heats is carried out. This is done by using the real gas equations but setting the coefficients of the Aungier-Redlich-Kwong equations to the appropriate values internally in the program to reproduce an ideal gas equation.

If i_fluid = 6 or 7 a calculation is done with a real gas with a constant real gas factor and a constant isentropic exponent.

Steam properties can be approximated with the Aungier-Redlich-Kwong equations using the values of the coefficients to model steam.

i_inbc = 0 then input values are total pressure, total temperature and swirl (r x cu, that is radius times circumferential component of the absolute velocity) on the input plane.

i_in_bc = 1 then input values are total pressure, total temperature, and absolute flow angle.

i_in_bc = 2 then input values are: total pressure, total temperature, and absolute circumferential flow velocity.

i_mass = 0 then the mass flow between each streamline is the same.

i_mass = 1 then the fractional mass between each streamline has to be input into the last line of the control file.

i_mass = 2 then the cumulative mass for each streamtube from the hub streamline is defined in the last line of the control file.

i_mix = 0 then no mixing model.

i_mix = 1 then a spanwise mixing model based on eddy diffusion across the streamlines will be used. Note that this requires a minimum of 9 streamlines (n_sl => 9).

i_ree = 0 then the equations as given in the paper of Casey and Roth (1984) are used, except that the dissipation term is set to zero and the blade force term is set to zero at a trailing edge and at a leading edge.

i_ree = 1 then the solution is as for i_ree = 0 but the dissipation term is not set to zero but the equations as given in the theory documentation are used.

i_ree = 2 then the dissipation term is not set to zero but the equations given in the paper of Casey and Roth (1984) are used.

i_ree = 3 then the velocity gradient in the radial equilibrium equation is reduced by the factor grad_ree given in section 5. This is useful for debugging difficult cases and the simulation becomes similar to a mean-line calculation with no gradient of meridional velocity across the span. If grad_ree = 1.0 then selecting i_ree = 3 has no effect.

The Wilkinson stability analysis for streamline curvature programs indicates that the streamline curvature damping term has to be reduced for long closely spaced quasi-orthogonals (high aspect ratio). For details, see the section on computational grid (Computational Grid).

If damp_sc = 0.0 then the program determines the value of the damping factor from the theory of Wilkinson, or uses an internally determined value that differs for each type of turbomachine — whichever is smaller.

Typically values of damp_sc between 0.05 and 0.25 are used, but lower values may be necessary for high aspect ratio quasi-orthogonals (as in end stages of steam turbines).

If the specified value of damp_sc is larger than the value predicted by the Wilkinson stability theory then the program automatically reduces the damping factor to a stable value.

If the program has convergence problems with an increasing error then the value is automatically reduced internally within the program during the convergence process.

Several different schemes for the damping are applied. The original scheme is obtained with a value of damp_sc between 1.0 and 1.25. New schemes which are more stable and robust in most cases can be obtained with the value of damp_sc between 0.0 and 0.25. A value between or 2.0 and 2.25 uses the original scheme with changed constants. A value between 3.0 and 3.25 gives a very robust scheme for highly staggered blades. The first digit then defines which scheme is used (0 - new scheme, 1 - original, 2 - modified original, 3 - new scheme for highly staggered blades) and the digits after the decimal point are the damping factor (damp_sc) itself.

There is no stability theory to define this, and a typical value used is 0.50, indicating that 50% of the new parameter together with 50% of its original value is used.

If damp_vl = 0.0 then 0.50 is used.

Recommended that this should be less than the actual value when converged because this avoids choking during the early streamline curvature iterations.

If simulations from a cold-start (with no restart file) fail to converge, it may be useful to modify this parameter because it strongly influences the start velocities.

Not used if the restart initial condition is used.

If cm_start = 0.0 then 0.25 is used for all calculations except radial turbines where 0.1 is used. The value of 0.25 is probably adequate for radial compressors only.

Typical value 0.01 (that is, 0.01%, which is 1 part in 10,000).

If tolerance_cm = 0.0 then 0.01% is used.

Note that if the meridional velocity is low at a certain point in the flow field, it may be necessary to use a higher value than this.

Note that the extremely low value of 0.01% does not imply that the solution is as accurate as this, but just provides confidence that convergence has really been achieved.

Typical value 0.001 (that is 0.001%, which is 1 part in 100000).

If error_max = 0.0 then 0.001% is used.

This line is only needed if i_mass = 1 or 2 on line 4 of the .con file, but for consistency, it is recommended to use it in all input files.

The number of values needed corresponds to the number of streamtubes, which is one less than the number of streamlines, so this line requires modification if the number of streamlines for the calculation is changed. The number of values and the values themselves need to change if the number of streamlines in the simulation is changed.

If i_mass = 1 then this line is interpreted as the relative mass flow for each streamtube from the hub, so the values of 1/(n_sl - 1) would give a uniform distribution. Each value may be specified as the relative mass flow for a streamtube compared to the sum of all values, so a value of 1.0 for each streamtube also leads to a uniform mass flow distribution.

If i_mass = 2 the cumulative values from the hub are specified, so the first value is 0.0 and the last value is 1.0.

It is recommended that the distribution is selected such that the central streamline always splits the flow into two regions of equal mass flow. The reason for this is that some correlations operate on the mean streamline.

The numbers may be specified on different lines.

Usually the geometry data is input in SI units (that is, all values are expected in m and not mm) and then this value is 1.0.

If the input geometry data comes from a CAD system then it may be in mm. In this case, the value of scale must be 0.001. Similarly the value can be adjusted to allow the geometry data to be input in other systems of units; for example, inches.

This scale factor only scales the data in the geometry input file and has no effect on other dimensions elsewhere; for example, it does not scale the reference diameter, which must be input with units of meters, in the aerodynamic file.

If you specify scale = 0.0, unity is used.

The actual value is not used by the program internally because it recounts the quasi-orthogonals as they are input. You can use this number to recognize a particular line of data in the input file.

This allows you to merge two different geometry files of, say, a rotor and stator, to a single stage geometry file without the need to renumber the quasi-orthogonals. In this case, the same number may appear more than once. In a similar way, a single quasi-orthogonal may be removed without the need for renumbering the lines.

r_hub may not be close to zero; adapt the grid if necessary to avoid small values of r_hub.

The aspect ratio of the quasi-orthogonal lines determines the stability of the solver. The lines should not be too closely spaced.

The end coordinates of the leading and trailing edges of blade rows should be included in the list of coordinates.

The end coordinates of the leading edge of a splitter vane should be included in the list of coordinates.

= 0 in duct regions.

The number of blades changes at a splitter blade leading edge in a compressor or at a splitter vane trailing edge in a turbine, and changes again for multiple splitters. This is the only information that the program has about the splitters, so the location of the splitter vane leading or trailing edge needs to be a quasi-orthogonal line in the input data.

= 1, a special case for duct stations. This indicates that there are 2 defining points (as for n_curve(i) = 2) but that no further information for this calculating station is provided in section 4 below, because it is already fully defined by the hub and casing points given in section 3.

= 2 for a linear calculating plane in which only the end points of the quasi-orthogonal are defined. In this case, similar information can be found in section 4.

> 2 for a non-linear or curved calculating plane.

The number of defining points can vary from station to station but n_curve(i) should typically be the same for all stations because streamline section data is usually available on a fixed number of spanwise sections.

Note that the hub and casing geometry information does not necessarily have to be the same as that defined in section 4. In this case the data in section 3 will be used to crop the blade row or to make a section through the blade information in section 4.

The program internally identifies which lines are leading and trailing edges from the changes of type of blade row, and which line is the leading edge of a splitter vane (by the change in blade number).

There must be at least two duct calculating stations upstream of the first blade row, and downstream of the last blade row.

A blade row must consist of at least two calculating stations (leading and trailing edge). Typically a radial impeller will have around 15 calculating stations, because this gives a 6° turn between each station and improves the calculation of the curvature terms.

There must be at least two blade calculating stations upstream and downstream of a splitter vane leading edge.

Other types of blade row may be defined at a later stage.

The angular coordinate (theta) is taken as positive in the clockwise direction of rotation and negative in the other direction.

In a duct region this angle may be zero.

This angle is not used by the program but helps to visualize the blade shape and may be useful for plots of the blade shape.

In a region where there is a splitter vane this angle is the blade camber angle of the main blade and not of the splitter, or it may be defined as the mean value for both blades.

In a duct region this thickness should be specified as zero.

In a region where there is a splitter vane this thickness is the mean thickness of the main blade and the splitter.

At leading and trailing edges the value supplied is not used by the program, but the calculating station is taken to be at the limit of the chord with zero thickness.

The thickness is not the thickness normal to the camber line.

If the value of zero is input then the program estimates the throat width from the blade angles and the blade thickness, taking into account that the throat is close to the leading edge for compressors and close to the trailing edge for turbines. For radial compressors this estimate is not particularly accurate and for cases close to choke you should provide more precise data here.

For axial compressors it is assumed that the blade has a circular arc camber line and the program includes an estimate of the throat area based on the geometrical relationships for circular arc blades.

For other blade types, you should specify the value.

If the value of zero is input then the program estimates the throat position from the blade angles and the blade thickness, taking into account that it is close to the leading edge for compressors and close to the trailing edge for turbines.

For radial compressors this estimate is not particularly accurate and for cases close to choke you should provide more precise data here.

For axial compressors it is assumed that the blade has a circular arc camber line and the program includes an estimate of the throat position based on the geometrical relationships for circular arc blades.

For other blade types, you should specify the value.

If the first value is non-zero, this is interpreted as the hub clearance.

If the last value (n_curve(i)) is non-zero, this is interpreted as the casing clearance.

In a variable stator vane with a hub clearance and a clearance gap for the shaft, it is possible to have non-zero values on both the hub and the tip and then both will be taken into account in the loss correlations.

Although the circumferential thickness of the blade row is already specified in section 3, the location of the quasi-orthogonal lines may not coincide with the location of normal maximum thickness, so this value must be specified separately.

If a value of zero is specified then the program searches for the normal maximum thickness of the blade, as given in section 3, and uses it to calculate the blade thickness.

Although the circumferential thickness of the blade row is already specified in section 3, this value must be specified separately.

If the value of zero is specified then the program estimates the trailing-edge thickness of the blade from the circumferential thickness given in section 3, and uses it.

Although the circumferential thickness of the blade row is already specified in section 3, this value must be specified separately.

If the value of zero is specified then the program estimates the leading-edge thickness of the blade from the circumferential thickness given in section 3, and uses it.

Although the chord of the blade row can be estimated from the meridional geometry and the blade angles, this is not exact so it is possible to specify this separately.

If a value of zero is specified then the program estimates the chord of the blade from the geometrical data in section 3, the meridional geometry, and the blade angles.

The blade row may be restaggered by specifying a change in the stagger angle in degrees. The change in angle only needs to be specified on the first of the streamline sections and is applied to all sections.

This is only valid for the restaggering of axial blade rows because it operates on the values of gamma_z_qo in section 4 by adding the change in stagger angle to those values.

A positive value increases the blade angles and the stagger, causing a more closed blade in a stator and a more open blade in a rotor (with negative gamma_z values).

If ref_mach < 3 then ref_mach is interpreted as ref_mach.

If ref_mach > 3 then ref_mach is interpreted as ref_u.

This should be less than the number of streamlines (n_inbc < n_sl).

if n_inbc = 1 then only a single value is input and this is taken to be on the mid-streamline. Note that a single value implies that the inlet conditions have a constant total pressure, constant total temperature, and constant swirl (rc_u) across the span at the inlet, whereby the value of the swirl is determined from the values specified (rcu, alpha, or cu).

A maximum of 25 points is allowed.

Typically 0.0001 to 0.001. See Streamline Curvature Throughflow Theory.

This has no function if i_mix = 0.

If i_mix = 1 and eddy = 0.0 then a value of eddy = 0.0005 will be used.

Typically, this line consists of a list of values such as 1.15, 1.10, 1.05, 1.00, 0.95, 0.90, and 0.85, indicating 115% down to 85% of the flow at the reference point.

If the value of i_flow is 1, the seven values listed above would be interpreted as seven different mass flows to be calculated. If the value of i_flow is 2, the values would be interpreted as different volume flows. If the value of i_flow is 0, the values would be interpreted as flow coefficients.

The flow scales linearly with the speed, so that when operating at the 100% flow point on the 90% speed line, the flow is reduced to 90% of the reference value.

A maximum of 11 points is allowed.

Typically, this section consists of a list of values such as 1.1, 1.0, 0.9, 0.8, 0.7, 0.65, 0.5.

These seven values would be interpreted as seven different speed lines to be calculated at 110%, 100%, 90%, etc., of the reference speed.

A maximum of 11 speeds is allowed.

i_loss = 0

No losses are specified and the calculation uses constant entropy. Note, however, that in order to retain similar structures for the correlations file, the lines described in sections 3 and 4 are still needed, but have no effect. This allows you to switch off the losses if necessary.

If i_loss = 1 then a user-defined variation of efficiency, loss coefficient, or dissipation coefficient across the span is specified at the spanwise positions given in section 4. The number of positions that are defined across the span and through the domain is given in section 3. Different values of the efficiency, loss coefficient, or dissipation coefficient can be specified for different quasi-orthogonal calculating stations and spanwise positions.

If i_loss = 1, then a value on the first quasi-orthogonal is always needed. This value remains constant until changed by the next quasi-orthogonal at which the loss is defined. If only a single value is specified across the span then this is applied on the mean streamline and each streamline has the same entropy rise as on the mean streamline. If a calculation with a constant efficiency for all streamlines is required then the efficiency needs to be specified as a constant on at least two points across the span.

If i_loss = 2 then built-in loss correlations are used according to the values given in sections 3 and 4.

If i_dev = 0, the flow follows the mean blade camber line given in the .geo file with no deviation. This option allows a mean S2 through-flow calculation to be carried out using the mean stream surface obtained from a series of S1 blade-to-blade calculations, provided that these are then given as the geometry of the camber surface in the .geo file. Note that, in order to retain similar structures for the correlations file, the lines described in sections 5 and 6 are still needed but have no effect.

If i_dev = 1 then data is specified in sections 5 and 6 to predict the variation of the flow angle at the blade outlet from the mean blade camber line across the span. The information can be either in the form of a deviation angle, a specified flow outlet angle, a specified slip factor, or a modification to the cosine rule, according to the type of blade row.

If i_dev = 2 then built-in correlations are used to determine the flow angle according to the type of blade row as specified in sections 5 and 6.

If i_ewb = 0 then no end-wall blockage is applied. Note that, in order to retain similar structures for the correlations file, the lines described in sections 7 and 8 are still needed, but have no effect.

If i_ewb = 1 then data is specified in sections 7 and 8 to predict the variation of the flow blockage for each quasi-orthogonal and for each streamline. A single value implies that a constant blockage value is applied in the whole calculation domain.

If i_ewb = 2 then builtin correlations are used to determine the blockage according to the type of blade row as specified in sections 7 and 8.

If k_loss = 1 then the small-scale static to static polytropic efficiency (eta_poly) is specified. This is used in such a way that the entropy always increases so that, for a local accelerating flow with a decrease in static enthalpy, it defines a turbine efficiency, and for a decelerating flow, it defines a compressor efficiency. This small-scale efficiency is applied in blade rows and in ducts. Note that specifying a specific value does not lead to an efficiency of the same value for the whole domain because this is dependent on both the local small-scale efficiency and the path of the process. For this, k_loss must be set to 4 or 7. See below.

If k_loss = 2 then the entropy loss coefficient (xsi) is specified. Note that the loss coefficient is the entropy loss coefficient with respect to outlet plane dynamic head for accelerating blade rows (turbine rotors and stators) and with respect to the inlet plane dynamic head for decelerating blade rows (compressor rotors and stators). An inlet guide vane for a compressor is therefore calculated as a turbine blade row, and an outlet guide vane for a turbine as a compressor blade row. Like the deviation, this value is only used at the trailing edge of a blade row.

If k_loss = 3 then the dissipation loss coefficient (cd) is specified to determine the losses from an integration of the loss production following the so-called u-cubed model of Denton. This is applied in blade rows and in ducts.

If k_loss = 4 to 9 then various forms of efficiency are used to specify the losses, as follows:

k_loss = 4 total-total polytropic efficiency

k_loss = 5 total-static polytropic efficiency

k_loss = 6 static-static polytropic efficiency

k_loss = 7 total-total isentropic efficiency

k_loss = 8 total-static isentropic efficiency

k_loss = 9 static-static isentropic efficiency

These are applied in blade rows and in duct regions.

In duct regions, these values are applied to the change from one quasi-orthogonal to the next. Clearly if the total enthalpy does not change, the use of 4 or 7, which give a total-total efficiency, is meaningless. In blade rows the full range is allowed, but options 6 and 9 are probably not very relevant. Use of 5 and 8 is not recommended because specifying a total-static efficiency that is too high may lead to a total-total efficiency that is greater than unity.

Option 4 or 7 is now recommended for rotor blade rows. This change forces the impeller total-total efficiency to be exactly the value that is specified. It is more robust, especially with real gas equations, because only the velocity triangle at the impeller outlet is involved in the estimate of the losses, and there is no need to use the real gas equations at each point in the stage when determining the entropy. It also has the advantage that the losses become exactly as specified, which was not the case with the option using a small-scale local polytropic efficiency.

If k_loss = 10 then the loss in a stator vane can be specified as a total pressure loss coefficient.

If k_loss = 11 then the performance of a stator vane can be specified as a static pressure rise coefficient.

These options have been programmed so that different blade rows may have different values of k_loss.

Radial Compressor Correlations

Axial Compressor Correlations

Axial Turbine Correlations

If k_dev = 1 then the deviation angle is specified (in degrees).

If k_dev = 2 then the relative outlet flow angle is specified (in degrees).

If k_dev = 3 then a slip factor is specified and applied on the mean line (dimensionless). Note that if no value is specified then the Wiesner slip factor is used. Only recommended for radial impellers and not for mixed flow stages, which should use deviation.

If k_dev = 4 then a slip factor is applied at each radius of the trailing edge (dimensionless). Note that if no value is specified then the Wiesner slip factor is used. Only recommended for radial impellers and not for mixed flow stages, which should use deviation.

If k_dev = 5, this is analogous to k_dev = 3, but the slip factor correlation of Xuwen Qiu is applied on the mean-line.

If k_dev = 6, this is analogous to k_dev = 4, but the slip factor correlation of Xuwen Qiu is applied on each streamline separately.

Radial Compressor Correlations

Axial Compressor Correlations

Axial Turbine Correlations

More General Correlations

i_ewb_type = 200

Recommended (202)

i_ewb_type = 201

Blockage calculated using the Miller Wright method adjusted to allow the blockage to vary through the blade row. The blockage value, however, is not used.

i_ewb_type = 202

Blockage calculated as in i_ewb_type = 201. The calculated values are used. The calculation leads to different values of blockage at all quasi-orthogonals and is based on a throughflow interpretation of the Miller Wright method.

i_ewb_type = 203

Blockage calculated and not used. The value of the blockage is calculated at trailing edges only and then remains constant to the next trailing edge. This is done to be consistent with SC90C blockage calculations, which are based on a ductflow interpretation of the Miller Wright method.