2001-10-05 20:27:01 +00:00
// IO360.cxx - a piston engine model currently for the IO360 engine fitted to the C172
// but with the potential to model other naturally aspirated piston engines
// given appropriate config input.
//
// Written by David Luff, started 2000.
// Based on code by Phil Schubert, started 1999.
2000-09-26 23:37:26 +00:00
//
// This program is free software; you can redistribute it and/or
// modify it under the terms of the GNU General Public License as
// published by the Free Software Foundation; either version 2 of the
// License, or (at your option) any later version.
//
// This program is distributed in the hope that it will be useful, but
// WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
// General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
2001-10-05 20:27:01 +00:00
// Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
2000-09-26 23:37:26 +00:00
2000-10-02 21:49:04 +00:00
# include <simgear/compiler.h>
2000-09-26 23:37:26 +00:00
# include <math.h>
2001-03-26 18:22:31 +00:00
# include STL_FSTREAM
# include STL_IOSTREAM
2000-09-26 23:37:26 +00:00
2003-05-20 11:29:06 +00:00
# include <Main/fg_props.hxx>
2001-03-26 18:22:31 +00:00
# include "IO360.hxx"
2003-05-20 11:29:06 +00:00
# include "ls_constants.h"
2001-03-26 18:22:31 +00:00
2000-09-26 23:37:26 +00:00
2001-02-02 20:55:41 +00:00
//*************************************************************************************
// Initialise the engine model
2000-10-16 20:01:23 +00:00
void FGNewEngine : : init ( double dt ) {
2000-09-26 23:37:26 +00:00
2001-02-02 20:55:41 +00:00
// These constants should probably be moved eventually
2000-09-26 23:37:26 +00:00
CONVERT_CUBIC_INCHES_TO_METERS_CUBED = 1.638706e-5 ;
2001-02-02 20:55:41 +00:00
CONVERT_HP_TO_WATTS = 745.6999 ;
// Properties of working fluids
2000-09-26 23:37:26 +00:00
Cp_air = 1005 ; // J/KgK
Cp_fuel = 1700 ; // J/KgK
calorific_value_fuel = 47.3e6 ; // W/Kg Note that this is only an approximate value
2001-03-01 16:32:29 +00:00
rho_fuel = 800 ; // kg/m^3 - an estimate for now
2000-09-26 23:37:26 +00:00
R_air = 287.3 ;
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
// environment inputs
p_amb_sea_level = 101325 ; // Pascals
// Control inputs - ARE THESE NEEDED HERE???
2001-02-02 20:55:41 +00:00
Throttle_Lever_Pos = 75 ;
Propeller_Lever_Pos = 75 ;
Mixture_Lever_Pos = 100 ;
2001-10-05 20:27:01 +00:00
//misc
IAS = 0 ;
2000-09-28 22:49:33 +00:00
time_step = dt ;
2000-09-26 23:37:26 +00:00
2001-10-05 20:27:01 +00:00
// Engine Specific Variables that should be read in from a config file
2001-09-19 22:58:29 +00:00
MaxHP = 200 ; //Lycoming IO360 -A-C-D series
// MaxHP = 180; //Current Lycoming IO360 ?
2000-09-26 23:37:26 +00:00
// displacement = 520; //Continental IO520-M
2001-02-02 20:55:41 +00:00
displacement = 360 ; //Lycoming IO360
displacement_SI = displacement * CONVERT_CUBIC_INCHES_TO_METERS_CUBED ;
2000-09-28 22:49:33 +00:00
engine_inertia = 0.2 ; //kgm^2 - value taken from a popular family saloon car engine - need to find an aeroengine value !!!!!
2001-09-19 22:58:29 +00:00
prop_inertia = 0.05 ; //kgm^2 - this value is a total guess - dcl
2001-10-05 20:27:01 +00:00
Max_Fuel_Flow = 130 ; // Units??? Do we need this variable any more??
// Engine specific variables that maybe should be read in from config but are pretty generic and won't vary much for a naturally aspirated piston engine.
Max_Manifold_Pressure = 28.50 ; //Inches Hg. An approximation - should be able to find it in the engine performance data
Min_Manifold_Pressure = 6.5 ; //Inches Hg. This is a guess corresponding to approx 0.24 bar MAP (7 in Hg) - need to find some proper data for this
Max_RPM = 2700 ;
Min_RPM = 600 ; //Recommended idle from Continental data sheet
Mag_Derate_Percent = 5 ;
2000-09-26 23:37:26 +00:00
Gear_Ratio = 1 ;
2001-10-05 20:27:01 +00:00
n_R = 2 ; // Number of crank revolutions per power cycle - 2 for a 4 stroke engine.
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
// Various bits of housekeeping describing the engines initial state.
2001-10-10 17:50:20 +00:00
running = false ;
2000-09-26 23:37:26 +00:00
cranking = false ;
2001-10-05 20:27:01 +00:00
crank_counter = false ;
2001-10-10 17:50:20 +00:00
starter = false ;
2000-09-26 23:37:26 +00:00
// Initialise Engine Variables used by this instance
2001-09-19 22:58:29 +00:00
if ( running )
RPM = 600 ;
else
RPM = 0 ;
2000-09-26 23:37:26 +00:00
Percentage_Power = 0 ;
2001-10-05 20:27:01 +00:00
Manifold_Pressure = 29.96 ; // Inches
2001-03-01 16:32:29 +00:00
Fuel_Flow_gals_hr = 0 ;
2001-10-05 20:27:01 +00:00
// Torque = 0;
2001-02-02 20:55:41 +00:00
Torque_SI = 0 ;
2002-02-08 17:27:38 +00:00
CHT = 298.0 ; //deg Kelvin
2001-10-05 20:27:01 +00:00
CHT_degF = ( CHT_degF * 1.8 ) - 459.67 ; //deg Fahrenheit
2000-09-26 23:37:26 +00:00
Mixture = 14 ;
Oil_Pressure = 0 ; // PSI
Oil_Temp = 85 ; // Deg C
2001-10-05 20:27:01 +00:00
current_oil_temp = 298.0 ; //deg Kelvin
/**** one of these is superfluous !!!!***/
2000-09-26 23:37:26 +00:00
HP = 0 ;
RPS = 0 ;
Torque_Imbalance = 0 ;
// Initialise Propellor Variables used by this instance
FGProp1_RPS = 0 ;
2001-10-05 20:27:01 +00:00
// Hardcode propellor for now - the following two should be read from config eventually
2001-03-22 16:27:16 +00:00
prop_diameter = 1.8 ; // meters
blade_angle = 23.0 ; // degrees
2000-09-26 23:37:26 +00:00
}
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
//*****************************************************************************
2000-09-26 23:37:26 +00:00
// update the engine model based on current control positions
2000-10-16 20:01:23 +00:00
void FGNewEngine : : update ( ) {
2001-03-22 16:27:16 +00:00
/*
// Hack for testing - should output every 5 seconds
static int count1 = 0 ;
if ( count1 = = 0 ) {
2001-03-22 23:42:16 +00:00
// cout << "P_atmos = " << p_amb << " T_atmos = " << T_amb << '\n';
// cout << "Manifold pressure = " << Manifold_Pressure << " True_Manifold_Pressure = " << True_Manifold_Pressure << '\n';
// cout << "p_amb_sea_level = " << p_amb_sea_level << '\n';
// cout << "equivalence_ratio = " << equivalence_ratio << '\n';
// cout << "combustion_efficiency = " << combustion_efficiency << '\n';
// cout << "AFR = " << 14.7 / equivalence_ratio << '\n';
// cout << "Mixture lever = " << Mixture_Lever_Pos << '\n';
// cout << "n = " << RPM << " rpm\n";
2001-09-19 22:58:29 +00:00
// cout << "T_amb = " << T_amb << '\n';
2001-10-05 20:27:01 +00:00
// cout << "running = " << running << '\n';
2003-03-20 10:43:02 +00:00
// cout << "fuel = " << fgGetFloat("/consumables/fuel/tank[0]/level-gal_us") << '\n';
2002-02-08 17:27:38 +00:00
// cout << "Percentage_Power = " << Percentage_Power << '\n';
// cout << "current_oil_temp = " << current_oil_temp << '\n';
2003-03-20 10:43:02 +00:00
// cout << "EGT = " << EGT << '\n';
2001-03-22 16:27:16 +00:00
}
count1 + + ;
2002-02-08 17:27:38 +00:00
if ( count1 = = 100 )
2001-03-22 16:27:16 +00:00
count1 = 0 ;
*/
2001-10-05 20:27:01 +00:00
// Check parameters that may alter the operating state of the engine.
// (spark, fuel, starter motor etc)
2000-09-26 23:37:26 +00:00
2001-09-19 22:58:29 +00:00
// Check for spark
2001-10-05 20:27:01 +00:00
bool Magneto_Left = false ;
bool Magneto_Right = false ;
2001-09-19 22:58:29 +00:00
// Magneto positions:
// 0 -> off
// 1 -> left only
// 2 -> right only
// 3 -> both
if ( mag_pos ! = 0 ) {
spark = true ;
} else {
spark = false ;
} // neglects battery voltage, master on switch, etc for now.
if ( ( mag_pos = = 1 ) | | ( mag_pos > 2 ) )
2001-10-05 20:27:01 +00:00
Magneto_Left = true ;
2001-09-19 22:58:29 +00:00
if ( mag_pos > 1 )
2001-10-05 20:27:01 +00:00
Magneto_Right = true ;
2001-09-19 22:58:29 +00:00
// crude check for fuel
if ( ( fgGetFloat ( " /consumables/fuel/tank[0]/level-gal_us " ) > 0 ) | | ( fgGetFloat ( " /consumables/fuel/tank[1]/level-gal_us " ) > 0 ) ) {
fuel = true ;
} else {
fuel = false ;
} // Need to make this better, eg position of fuel selector switch.
// Check if we are turning the starter motor
2001-10-10 17:50:20 +00:00
if ( cranking ! = starter ) {
2001-09-19 22:58:29 +00:00
// This check saves .../cranking from getting updated every loop - they only update when changed.
2001-10-10 17:50:20 +00:00
cranking = starter ;
crank_counter = 0 ;
2001-09-19 22:58:29 +00:00
}
// Note that although /engines/engine[0]/starter and /engines/engine[0]/cranking might appear to be duplication it is
// not since the starter may be engaged with the battery voltage too low for cranking to occur (or perhaps the master
// switch just left off) and the sound manager will read .../cranking to determine wether to play a cranking sound.
// For now though none of that is implemented so cranking can be set equal to .../starter without further checks.
2000-09-26 23:37:26 +00:00
2001-10-05 20:27:01 +00:00
// int Alternate_Air_Pos =0; // Off = 0. Reduces power by 3 % for same throttle setting
2001-09-19 22:58:29 +00:00
// DCL - don't know what this Alternate_Air_Pos is - this is a leftover from the Schubert code.
//Check mode of engine operation
if ( cranking ) {
2001-10-05 20:27:01 +00:00
crank_counter + + ;
2001-09-19 22:58:29 +00:00
if ( RPM < = 480 ) {
RPM + = 100 ;
if ( RPM > 480 )
RPM = 480 ;
} else {
// consider making a horrible noise if the starter is engaged with the engine running
}
}
2001-10-05 20:27:01 +00:00
if ( ( ! running ) & & ( spark ) & & ( fuel ) & & ( crank_counter > 120 ) ) {
2001-09-19 22:58:29 +00:00
// start the engine if revs high enough
if ( RPM > 450 ) {
// For now just instantaneously start but later we should maybe crank for a bit
running = true ;
2001-10-05 20:27:01 +00:00
// RPM = 600;
2001-09-19 22:58:29 +00:00
}
}
if ( ( running ) & & ( ( ! spark ) | | ( ! fuel ) ) ) {
// Cut the engine
// note that we only cut the power - the engine may continue to spin if the prop is in a moving airstream
running = false ;
}
2000-09-26 23:37:26 +00:00
2001-10-05 20:27:01 +00:00
// Now we've ascertained whether the engine is running or not we can start to do the engine calculations 'proper'
2001-03-22 16:27:16 +00:00
// Calculate Sea Level Manifold Pressure
Manifold_Pressure = Calc_Manifold_Pressure ( Throttle_Lever_Pos , Max_Manifold_Pressure , Min_Manifold_Pressure ) ;
2000-09-26 23:37:26 +00:00
// cout << "manifold pressure = " << Manifold_Pressure << endl;
//Then find the actual manifold pressure (the calculated one is the sea level pressure)
True_Manifold_Pressure = Manifold_Pressure * p_amb / p_amb_sea_level ;
2001-10-05 20:27:01 +00:00
//Do the fuel flow calculations
Calc_Fuel_Flow_Gals_Hr ( ) ;
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
//Calculate engine power
Calc_Percentage_Power ( Magneto_Left , Magneto_Right ) ;
HP = Percentage_Power * MaxHP / 100.0 ;
Power_SI = HP * CONVERT_HP_TO_WATTS ;
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
// FMEP calculation. For now we will just use this during motored operation.
// Eventually we will calculate IMEP and use the FMEP all the time to give BMEP (maybe!)
if ( ! running ) {
// This FMEP data is from the Patton paper, assumes fully warm conditions.
FMEP = 1e-12 * pow ( RPM , 4 ) - 1e-8 * pow ( RPM , 3 ) + 5e-5 * pow ( RPM , 2 ) - 0.0722 * RPM + 154.85 ;
// Gives FMEP in kPa - now convert to Pa
FMEP * = 1000 ;
} else {
FMEP = 0.0 ;
}
// Is this total FMEP or friction FMEP ???
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
Torque_FMEP = ( FMEP * displacement_SI ) / ( 2.0 * LS_PI * n_R ) ;
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
// Calculate Engine Torque. Check for div by zero since percentage power correlation does not guarantee zero power at zero rpm.
// However this is problematical since there is a resistance to movement even at rest
// Ie this is a dynamics equation not a statics one. This can be solved by going over to MEP based torque calculations.
if ( RPM = = 0 ) {
Torque_SI = 0 - Torque_FMEP ;
}
else {
Torque_SI = ( ( Power_SI * 60.0 ) / ( 2.0 * LS_PI * RPM ) ) - Torque_FMEP ; //Torque = power / angular velocity
// cout << Torque << " Nm\n";
}
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
//Calculate Exhaust gas temperature
2002-02-08 17:27:38 +00:00
if ( running )
Calc_EGT ( ) ;
else
EGT = 298.0 ;
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
// Calculate Cylinder Head Temperature
2002-02-08 17:27:38 +00:00
Calc_CHT ( ) ;
2001-10-05 20:27:01 +00:00
// Calculate oil temperature
current_oil_temp = Calc_Oil_Temp ( current_oil_temp ) ;
// Calculate Oil Pressure
Oil_Pressure = Calc_Oil_Press ( Oil_Temp , RPM ) ;
// Now do the Propeller Calculations
Do_Prop_Calcs ( ) ;
// Now do the engine - prop torque balance to calculate final RPM
//Calculate new RPM from torque balance and inertia.
Torque_Imbalance = Torque_SI - prop_torque ; //This gives a +ve value when the engine torque exeeds the prop torque
// (Engine torque is +ve when it acts in the direction of engine revolution, prop torque is +ve when it opposes the direction of engine revolution)
angular_acceleration = Torque_Imbalance / ( engine_inertia + prop_inertia ) ;
angular_velocity_SI + = ( angular_acceleration * time_step ) ;
// Don't let the engine go into reverse
if ( angular_velocity_SI < 0 )
angular_velocity_SI = 0 ;
RPM = ( angular_velocity_SI * 60 ) / ( 2.0 * LS_PI ) ;
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
// And finally a last check on the engine state after doing the torque balance with the prop - have we stalled?
if ( running ) {
//Check if we have stalled the engine
if ( RPM = = 0 ) {
running = false ;
} else if ( ( RPM < = 480 ) & & ( cranking ) ) {
//Make sure the engine noise dosn't play if the engine won't start due to eg mixture lever pulled out.
running = false ;
2002-02-08 17:27:38 +00:00
EGT = 298.0 ;
2001-10-05 20:27:01 +00:00
}
2001-02-02 20:55:41 +00:00
}
2002-02-08 17:27:38 +00:00
// And finally, do any unit conversions from internal units to output units
EGT_degF = ( EGT * 1.8 ) - 459.67 ;
CHT_degF = ( CHT * 1.8 ) - 459.67 ;
2001-10-05 20:27:01 +00:00
}
2001-02-02 20:55:41 +00:00
2001-10-05 20:27:01 +00:00
//*****************************************************************************************************
2001-09-19 22:58:29 +00:00
2001-10-05 20:27:01 +00:00
// FGNewEngine member functions
2001-09-19 22:58:29 +00:00
2001-10-05 20:27:01 +00:00
////////////////////////////////////////////////////////////////////////////////////////////
// Return the combustion efficiency as a function of equivalence ratio
//
// Combustion efficiency values from Heywood,
// "Internal Combustion Engine Fundamentals", ISBN 0-07-100499-8
////////////////////////////////////////////////////////////////////////////////////////////
float FGNewEngine : : Lookup_Combustion_Efficiency ( float thi_actual )
{
const int NUM_ELEMENTS = 11 ;
float thi [ NUM_ELEMENTS ] = { 0.0 , 0.9 , 1.0 , 1.05 , 1.1 , 1.15 , 1.2 , 1.3 , 1.4 , 1.5 , 1.6 } ; //array of equivalence ratio values
float neta_comb [ NUM_ELEMENTS ] = { 0.98 , 0.98 , 0.97 , 0.95 , 0.9 , 0.85 , 0.79 , 0.7 , 0.63 , 0.57 , 0.525 } ; //corresponding array of combustion efficiency values
float neta_comb_actual = 0.0f ;
float factor ;
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
int i ;
int j = NUM_ELEMENTS ; //This must be equal to the number of elements in the lookup table arrays
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
for ( i = 0 ; i < j ; i + + )
{
if ( i = = ( j - 1 ) ) {
// Assume linear extrapolation of the slope between the last two points beyond the last point
float dydx = ( neta_comb [ i ] - neta_comb [ i - 1 ] ) / ( thi [ i ] - thi [ i - 1 ] ) ;
neta_comb_actual = neta_comb [ i ] + dydx * ( thi_actual - thi [ i ] ) ;
return neta_comb_actual ;
}
if ( thi_actual = = thi [ i ] ) {
neta_comb_actual = neta_comb [ i ] ;
return neta_comb_actual ;
}
if ( ( thi_actual > thi [ i ] ) & & ( thi_actual < thi [ i + 1 ] ) ) {
//do linear interpolation between the two points
factor = ( thi_actual - thi [ i ] ) / ( thi [ i + 1 ] - thi [ i ] ) ;
neta_comb_actual = ( factor * ( neta_comb [ i + 1 ] - neta_comb [ i ] ) ) + neta_comb [ i ] ;
return neta_comb_actual ;
}
2001-02-02 20:55:41 +00:00
}
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
//if we get here something has gone badly wrong
2003-03-20 10:43:02 +00:00
SG_LOG ( SG_AIRCRAFT , SG_ALERT , " error in FGNewEngine::Lookup_Combustion_Efficiency " ) ;
2001-10-05 20:27:01 +00:00
return neta_comb_actual ;
}
2000-09-26 23:37:26 +00:00
2001-10-05 20:27:01 +00:00
////////////////////////////////////////////////////////////////////////////////////////////
// Return the percentage of best mixture power available at a given mixture strength
//
// Based on data from "Technical Considerations for Catalysts for the European Market"
// by H S Gandi, published 1988 by IMechE
//
// Note that currently no attempt is made to set a lean limit on stable combustion
////////////////////////////////////////////////////////////////////////////////////////////
float FGNewEngine : : Power_Mixture_Correlation ( float thi_actual )
{
float AFR_actual = 14.7 / thi_actual ;
// thi and thi_actual are equivalence ratio
const int NUM_ELEMENTS = 13 ;
// The lookup table is in AFR because the source data was. I added the two end elements to make sure we are almost always in it.
float AFR [ NUM_ELEMENTS ] = { ( 14.7 / 1.6 ) , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , ( 14.7 / 0.6 ) } ; //array of equivalence ratio values
float mixPerPow [ NUM_ELEMENTS ] = { 78 , 86 , 93.5 , 98 , 100 , 99 , 96.4 , 92.5 , 88 , 83 , 78.5 , 74 , 58 } ; //corresponding array of combustion efficiency values
float mixPerPow_actual = 0.0f ;
float factor ;
float dydx ;
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
int i ;
int j = NUM_ELEMENTS ; //This must be equal to the number of elements in the lookup table arrays
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
for ( i = 0 ; i < j ; i + + )
{
if ( i = = ( j - 1 ) ) {
// Assume linear extrapolation of the slope between the last two points beyond the last point
dydx = ( mixPerPow [ i ] - mixPerPow [ i - 1 ] ) / ( AFR [ i ] - AFR [ i - 1 ] ) ;
mixPerPow_actual = mixPerPow [ i ] + dydx * ( AFR_actual - AFR [ i ] ) ;
return mixPerPow_actual ;
}
if ( ( i = = 0 ) & & ( AFR_actual < AFR [ i ] ) ) {
// Assume linear extrapolation of the slope between the first two points for points before the first point
dydx = ( mixPerPow [ i ] - mixPerPow [ i - 1 ] ) / ( AFR [ i ] - AFR [ i - 1 ] ) ;
mixPerPow_actual = mixPerPow [ i ] + dydx * ( AFR_actual - AFR [ i ] ) ;
return mixPerPow_actual ;
}
if ( AFR_actual = = AFR [ i ] ) {
mixPerPow_actual = mixPerPow [ i ] ;
return mixPerPow_actual ;
}
if ( ( AFR_actual > AFR [ i ] ) & & ( AFR_actual < AFR [ i + 1 ] ) ) {
//do linear interpolation between the two points
factor = ( AFR_actual - AFR [ i ] ) / ( AFR [ i + 1 ] - AFR [ i ] ) ;
mixPerPow_actual = ( factor * ( mixPerPow [ i + 1 ] - mixPerPow [ i ] ) ) + mixPerPow [ i ] ;
return mixPerPow_actual ;
}
}
2000-11-02 17:01:09 +00:00
2001-10-05 20:27:01 +00:00
//if we get here something has gone badly wrong
2003-03-20 10:43:02 +00:00
SG_LOG ( SG_AIRCRAFT , SG_ALERT , " error in FGNewEngine::Power_Mixture_Correlation " ) ;
2001-10-05 20:27:01 +00:00
return mixPerPow_actual ;
}
2000-09-26 23:37:26 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
// Calculate Cylinder Head Temperature
2001-10-05 20:27:01 +00:00
// Crudely models the cylinder head as an arbitary lump of arbitary size and area with one third of combustion energy
// as heat input and heat output as a function of airspeed and temperature. Could be improved!!!
2002-02-08 17:27:38 +00:00
void FGNewEngine : : Calc_CHT ( )
2001-10-05 20:27:01 +00:00
{
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
float h1 = - 95.0 ; //co-efficient for free convection
float h2 = - 3.95 ; //co-efficient for forced convection
float h3 = - 0.05 ; //co-efficient for forced convection due to prop backwash
float v_apparent ; //air velocity over cylinder head in m/s
float v_dot_cooling_air ;
float m_dot_cooling_air ;
float temperature_difference ;
float arbitary_area = 1.0 ;
float dqdt_from_combustion ;
float dqdt_forced ; //Rate of energy transfer to/from cylinder head due to forced convection (Joules) (sign convention: to cylinder head is +ve)
float dqdt_free ; //Rate of energy transfer to/from cylinder head due to free convection (Joules) (sign convention: to cylinder head is +ve)
float dqdt_cylinder_head ; //Overall energy change in cylinder head
float CpCylinderHead = 800.0 ; //FIXME - this is a guess - I need to look up the correct value
float MassCylinderHead = 8.0 ; //Kg - this is a guess - it dosn't have to be absolutely accurate but can be varied to alter the warm-up rate
float HeatCapacityCylinderHead ;
float dCHTdt ;
2001-10-05 20:27:01 +00:00
// The above values are hardwired to give reasonable results for an IO360 (C172 engine)
// Now adjust to try to compensate for arbitary sized engines - this is currently untested
arbitary_area * = ( MaxHP / 180.0 ) ;
MassCylinderHead * = ( MaxHP / 180.0 ) ;
2001-02-02 20:55:41 +00:00
temperature_difference = CHT - T_amb ;
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
v_apparent = IAS * 0.5144444 ; //convert from knots to m/s
v_dot_cooling_air = arbitary_area * v_apparent ;
m_dot_cooling_air = v_dot_cooling_air * rho_air ;
//Calculate rate of energy transfer to cylinder head from combustion
dqdt_from_combustion = m_dot_fuel * calorific_value_fuel * combustion_efficiency * 0.33 ;
2001-03-22 16:27:16 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
//Calculate rate of energy transfer from cylinder head due to cooling NOTE is calculated as rate to but negative
dqdt_forced = ( h2 * m_dot_cooling_air * temperature_difference ) + ( h3 * RPM * temperature_difference ) ;
dqdt_free = h1 * temperature_difference ;
2001-03-22 16:27:16 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
//Calculate net rate of energy transfer to or from cylinder head
dqdt_cylinder_head = dqdt_from_combustion + dqdt_forced + dqdt_free ;
2001-03-22 16:27:16 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
HeatCapacityCylinderHead = CpCylinderHead * MassCylinderHead ;
2001-03-22 16:27:16 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
dCHTdt = dqdt_cylinder_head / HeatCapacityCylinderHead ;
2001-03-22 16:27:16 +00:00
This is a somewhat rough first attempt at modelling cylinder head
temperature. The cylinder head is assumed to be at uniform
temperature. Obviously this is incorrect, but it simplifies things a
lot, and we're just looking for the behaviour of CHT to be correct.
Energy transfer to the cylinder head is assumed to be one third of the
energy released by combustion at all conditions. This is a reasonable
estimate, although obviously in real life it varies with different
conditions and possibly with CHT itself. I've split energy transfer
from the cylinder head into 2 terms - free convection - ie convection
to stationary air, and forced convection, ie convection into flowing
air. The basic free convection equation is: dqdt = -hAdT Since we
don't know A and are going to set h quite arbitarily anyway I've
knocked A out and just wrapped it up in h - the only real significance
is that the units of h will be different but that dosn't really matter
to us anyway. In addition, we have the problem that the prop model
I'm currently using dosn't model the backwash from the prop which will
add to the velocity of the cooling air when the prop is turning, so
I've added an extra term to try and cope with this.
In real life, forced convection equations are genarally empirically
derived, and are quite complicated and generally contain such things
as the Reynolds and Nusselt numbers to various powers. The best
course of action would probably to find an empirical correlation from
the literature for a similar situation and try and get it to fit well.
However, for now I am using my own made up very simple correlation
for the energy transfer from the cylinder head:
dqdt = -(h1.dT) -(h2.m_dot.dT) -(h3.rpm.dT)
where dT is the temperature different between the cylinder head and
the surrounding air, m_dot is the mass flow rate of cooling air
through an arbitary volume, rpm is the engine speed in rpm (this is
the backwash term), and h1, h2, h3 are co-efficients which we can play
with to attempt to get the CHT behaviour to match real life.
In order to change the values of CHT that the engine settles down at
at various conditions, have a play with h1, h2 and h3. In order to
change the rate of heating/cooling without affecting equilibrium
values alter the cylinder head mass, which is really quite arbitary.
Bear in mind that altering h1, h2 and h3 will also alter the rate of
heating or cooling as well as equilibrium values, but altering the
cylinder head mass will only alter the rate. It would I suppose be
better to read the values from file to avoid the necessity for
re-compilation every time I change them.
2000-10-27 21:33:07 +00:00
CHT + = ( dCHTdt * time_step ) ;
2001-10-05 20:27:01 +00:00
}
// Calculate exhaust gas temperature
void FGNewEngine : : Calc_EGT ( )
{
combustion_efficiency = Lookup_Combustion_Efficiency ( equivalence_ratio ) ; //The combustion efficiency basically tells us what proportion of the fuels calorific value is released
//now calculate energy release to exhaust
//We will assume a three way split of fuel energy between useful work, the coolant system and the exhaust system
//This is a reasonable first suck of the thumb estimate for a water cooled automotive engine - whether it holds for an air cooled aero engine is probably open to question
//Regardless - it won't affect the variation of EGT with mixture, and we can always put a multiplier on EGT to get a reasonable peak value.
enthalpy_exhaust = m_dot_fuel * calorific_value_fuel * combustion_efficiency * 0.33 ;
heat_capacity_exhaust = ( Cp_air * m_dot_air ) + ( Cp_fuel * m_dot_fuel ) ;
delta_T_exhaust = enthalpy_exhaust / heat_capacity_exhaust ;
EGT = T_amb + delta_T_exhaust ;
//The above gives the exhaust temperature immediately prior to leaving the combustion chamber
//Now derate to give a more realistic figure as measured downstream
//For now we will aim for a peak of around 400 degC (750 degF)
EGT * = 0.444 + ( ( 0.544 - 0.444 ) * Percentage_Power / 100.0 ) ;
}
// Calculate Manifold Pressure based on Throttle lever Position
float FGNewEngine : : Calc_Manifold_Pressure ( float LeverPosn , float MaxMan , float MinMan )
{
float Inches ;
//Note that setting the manifold pressure as a function of lever position only is not strictly accurate
//MAP is also a function of engine speed. (and ambient pressure if we are going for an actual MAP model)
Inches = MinMan + ( LeverPosn * ( MaxMan - MinMan ) / 100 ) ;
//allow for idle bypass valve or slightly open throttle stop
if ( Inches < MinMan )
Inches = MinMan ;
//Check for stopped engine (crudest way of compensating for engine speed)
if ( RPM = = 0 )
Inches = 29.92 ;
return Inches ;
}
// Calculate fuel flow in gals/hr
void FGNewEngine : : Calc_Fuel_Flow_Gals_Hr ( )
{
//DCL - calculate mass air flow into engine based on speed and load - separate this out into a function eventually
//t_amb is actual temperature calculated from altitude
//calculate density from ideal gas equation
rho_air = p_amb / ( R_air * T_amb ) ;
rho_air_manifold = rho_air * Manifold_Pressure / 29.6 ; //This is a bit of a roundabout way of calculating this but it works !! If we put manifold pressure into SI units we could do it simpler.
//calculate ideal engine volume inducted per second
swept_volume = ( displacement_SI * ( RPM / 60 ) ) / 2 ; //This equation is only valid for a four stroke engine
//calculate volumetric efficiency - for now we will just use 0.8, but actually it is a function of engine speed and the exhaust to manifold pressure ratio
//Note that this is cylinder vol eff - the throttle drop is already accounted for in the MAP calculation
volumetric_efficiency = 0.8 ;
//Now use volumetric efficiency to calculate actual air volume inducted per second
v_dot_air = swept_volume * volumetric_efficiency ;
//Now calculate mass flow rate of air into engine
m_dot_air = v_dot_air * rho_air_manifold ;
//**************
//DCL - now calculate fuel flow into engine based on air flow and mixture lever position
//assume lever runs from no flow at fully out to thi = 1.3 at fully in at sea level
//also assume that the injector linkage is ideal - hence the set mixture is maintained at a given altitude throughout the speed and load range
thi_sea_level = 1.3 * ( Mixture_Lever_Pos / 100.0 ) ;
equivalence_ratio = thi_sea_level * p_amb_sea_level / p_amb ; //ie as we go higher the mixture gets richer for a given lever position
m_dot_fuel = m_dot_air / 14.7 * equivalence_ratio ;
Fuel_Flow_gals_hr = ( m_dot_fuel / rho_fuel ) * 264.172 * 3600.0 ; // Note this assumes US gallons
}
// Calculate the percentage of maximum rated power delivered as a function of Manifold pressure multiplied by engine speed (rpm)
// This is not necessarilly the best approach but seems to work for now.
// May well need tweaking at the bottom end if the prop model is changed.
void FGNewEngine : : Calc_Percentage_Power ( bool mag_left , bool mag_right )
{
// For a given Manifold Pressure and RPM calculate the % Power
// Multiply Manifold Pressure by RPM
float ManXRPM = True_Manifold_Pressure * RPM ;
/*
// Phil's %power correlation
// Calculate % Power
Percentage_Power = ( + 7E-09 * ManXRPM * ManXRPM ) + ( + 7E-04 * ManXRPM ) - 0.1218 ;
// cout << Percentage_Power << "%" << "\t";
*/
// DCL %power correlation - basically Phil's correlation modified to give slighty less power at the low end
// might need some adjustment as the prop model is adjusted
// My aim is to match the prop model and engine model at the low end to give the manufacturer's recommended idle speed with the throttle closed - 600rpm for the Continental IO520
// Calculate % Power for Nev's prop model
//Percentage_Power = (+ 6E-09 * ManXRPM * ManXRPM) + ( + 8E-04 * ManXRPM) - 1.8524;
// Calculate %power for DCL prop model
Percentage_Power = ( 7e-9 * ManXRPM * ManXRPM ) + ( 7e-4 * ManXRPM ) - 1.0 ;
// Power de-rating for altitude has been removed now that we are basing the power
// on the actual manifold pressure, which takes air pressure into account. However - this fails to
// take the temperature into account - this is TODO.
// Adjust power for temperature - this is temporary until the power is done as a function of mass flow rate induced
// Adjust for Temperature - Temperature above Standard decrease
// power by 7/120 % per degree F increase, and incease power for
// temps below at the same ratio
float T_amb_degF = ( T_amb * 1.8 ) - 459.67 ;
float T_amb_sea_lev_degF = ( 288 * 1.8 ) - 459.67 ;
Percentage_Power = Percentage_Power + ( ( T_amb_sea_lev_degF - T_amb_degF ) * 7 / 120 ) ;
//DCL - now adjust power to compensate for mixture
Percentage_of_best_power_mixture_power = Power_Mixture_Correlation ( equivalence_ratio ) ;
Percentage_Power = Percentage_Power * Percentage_of_best_power_mixture_power / 100.0 ;
// Now Derate engine for the effects of Bad/Switched off magnetos
//if (Magneto_Left == 0 && Magneto_Right == 0) {
if ( ! running ) {
// cout << "Both OFF\n";
Percentage_Power = 0 ;
} else if ( mag_left & & mag_right ) {
// cout << "Both On ";
} else if ( mag_left = = 0 | | mag_right = = 0 ) {
// cout << "1 Magneto Failed ";
Percentage_Power = Percentage_Power * ( ( 100.0 - Mag_Derate_Percent ) / 100.0 ) ;
// cout << FGEng1_Percentage_Power << "%" << "\t";
}
/*
//DCL - stall the engine if RPM drops below 450 - this is possible if mixture lever is pulled right out
//This is a kludge that I should eliminate by adding a total fmep estimation.
if ( RPM < 450 )
Percentage_Power = 0 ;
*/
if ( Percentage_Power < 0 )
Percentage_Power = 0 ;
}
// Calculate Oil Temperature in degrees Kelvin
float FGNewEngine : : Calc_Oil_Temp ( float oil_temp )
{
float idle_percentage_power = 2.3 ; // approximately
float target_oil_temp ; // Steady state oil temp at the current engine conditions
float time_constant ; // The time constant for the differential equation
if ( running ) {
target_oil_temp = 363 ;
time_constant = 500 ; // Time constant for engine-on idling.
if ( Percentage_Power > idle_percentage_power ) {
time_constant / = ( ( Percentage_Power / idle_percentage_power ) / 10.0 ) ; // adjust for power
}
} else {
target_oil_temp = 298 ;
time_constant = 1000 ; // Time constant for engine-off; reflects the fact that oil is no longer getting circulated
}
float dOilTempdt = ( target_oil_temp - oil_temp ) / time_constant ;
oil_temp + = ( dOilTempdt * time_step ) ;
return ( oil_temp ) ;
}
// Calculate Oil Pressure
float FGNewEngine : : Calc_Oil_Press ( float Oil_Temp , float Engine_RPM )
{
float Oil_Pressure = 0 ; //PSI
float Oil_Press_Relief_Valve = 60 ; //PSI
float Oil_Press_RPM_Max = 1800 ;
float Design_Oil_Temp = 85 ; //Celsius
float Oil_Viscosity_Index = 0.25 ; // PSI/Deg C
// float Temp_Deviation = 0; // Deg C
Oil_Pressure = ( Oil_Press_Relief_Valve / Oil_Press_RPM_Max ) * Engine_RPM ;
// Pressure relief valve opens at Oil_Press_Relief_Valve PSI setting
if ( Oil_Pressure > = Oil_Press_Relief_Valve ) {
Oil_Pressure = Oil_Press_Relief_Valve ;
}
// Now adjust pressure according to Temp which affects the viscosity
Oil_Pressure + = ( Design_Oil_Temp - Oil_Temp ) * Oil_Viscosity_Index ;
return Oil_Pressure ;
}
// Propeller calculations.
void FGNewEngine : : Do_Prop_Calcs ( )
{
float Gear_Ratio = 1.0 ;
float forward_velocity ; // m/s
float prop_power_consumed_SI ; // Watts
double J ; // advance ratio - dimensionless
double Cp_20 ; // coefficient of power for 20 degree blade angle
double Cp_25 ; // coefficient of power for 25 degree blade angle
double Cp ; // Our actual coefficient of power
double neta_prop_20 ;
double neta_prop_25 ;
double neta_prop ; // prop efficiency
FGProp1_RPS = RPM * Gear_Ratio / 60.0 ;
2001-09-19 22:58:29 +00:00
angular_velocity_SI = 2.0 * LS_PI * RPM / 60.0 ;
2001-03-22 16:27:16 +00:00
forward_velocity = IAS * 0.514444444444 ; // Convert to m/s
if ( FGProp1_RPS = = 0 )
J = 0 ;
else
J = forward_velocity / ( FGProp1_RPS * prop_diameter ) ;
//cout << "advance_ratio = " << J << '\n';
//Cp correlations based on data from McCormick
Cp_20 = 0.0342 * J * J * J * J - 0.1102 * J * J * J + 0.0365 * J * J - 0.0133 * J + 0.064 ;
Cp_25 = 0.0119 * J * J * J * J - 0.0652 * J * J * J + 0.018 * J * J - 0.0077 * J + 0.0921 ;
//cout << "Cp_20 = " << Cp_20 << '\n';
//cout << "Cp_25 = " << Cp_25 << '\n';
//Assume that the blade angle is between 20 and 25 deg - reasonable for fixed pitch prop but won't hold for variable one !!!
Cp = Cp_20 + ( ( Cp_25 - Cp_20 ) * ( ( blade_angle - 20 ) / ( 25 - 20 ) ) ) ;
//cout << "Cp = " << Cp << '\n';
//cout << "RPM = " << RPM << '\n';
//cout << "angular_velocity_SI = " << angular_velocity_SI << '\n';
2002-01-10 12:52:39 +00:00
prop_power_consumed_SI = Cp * rho_air * pow ( FGProp1_RPS , 3.0f ) * pow ( float ( prop_diameter ) , 5.0f ) ;
2001-03-22 16:27:16 +00:00
//cout << "prop HP consumed = " << prop_power_consumed_SI / 745.699 << '\n';
if ( angular_velocity_SI = = 0 )
prop_torque = 0 ;
2001-09-19 22:58:29 +00:00
// However this can give problems - if rpm == 0 but forward velocity increases the prop should be able to generate a torque to start the engine spinning
// Unlikely to happen in practice - but I suppose someone could move the lever of a stopped large piston engine from feathered to windmilling.
// This *does* give problems - if the plane is put into a steep climb whilst windmilling the engine friction will eventually stop it spinning.
// When put back into a dive it never starts re-spinning again. Although it is unlikely that anyone would do this in real life, they might well do it in a sim!!!
2001-03-22 16:27:16 +00:00
else
prop_torque = prop_power_consumed_SI / angular_velocity_SI ;
// calculate neta_prop here
neta_prop_20 = 0.1328 * J * J * J * J - 1.3073 * J * J * J + 0.3525 * J * J + 1.5591 * J + 0.0007 ;
neta_prop_25 = - 0.3121 * J * J * J * J + 0.4234 * J * J * J - 0.7686 * J * J + 1.5237 * J - 0.0004 ;
neta_prop = neta_prop_20 + ( ( neta_prop_25 - neta_prop_20 ) * ( ( blade_angle - 20 ) / ( 25 - 20 ) ) ) ;
2001-09-19 22:58:29 +00:00
// Check for zero forward velocity to avoid divide by zero
2001-03-22 16:27:16 +00:00
if ( forward_velocity < 0.0001 )
prop_thrust = 0.0 ;
2001-09-19 22:58:29 +00:00
// I don't see how this works - how can the plane possibly start from rest ???
// Hmmmm - it works because the forward_velocity at present never drops below about 0.03 even at rest
// We can't really rely on this in the future - needs fixing !!!!
// The problem is that we're doing this calculation backwards - we're working out the thrust from the power consumed and the velocity, which becomes invalid as velocity goes to zero.
// It would be far more natural to work out the power consumed from the thrust - FIXME!!!!!.
2001-03-22 16:27:16 +00:00
else
prop_thrust = neta_prop * prop_power_consumed_SI / forward_velocity ; //TODO - rename forward_velocity to IAS_SI
//cout << "prop_thrust = " << prop_thrust << '\n';
2000-09-26 23:37:26 +00:00
}