a way to calculate torque/power from engine specifications?

[pik_d]

i've had the crazy idea of writing a (c++) program which would make a pretty display of HP and torque based on some form of input.

the most obvious way i thought of was to input either the hp or torque at certain intervals along the powercurve. but i dont really like that idea, to much to input to make it look smooth.

my next idea was to find some calculation which would take "SOMETHING" and do some nifty calculations and give power/torque numbers. yay.

now, for all you guru's, is there a way to take bore, stroke, cylinder numbers, compression ratio, etc, and make a ROUGH guess of horsepower and torque?

i looked around a bit, and found something called RAChp (royal automotive club, from england, also called taxable horsepower) which is this equation:

D^2 * (n*2.5)

where D is the bore in inches, and n is the number of cylinders.

now, this is obviously NOT going to be close at all, but it's all i've found. (by that, an SRT-4 gets 136hp. )

so... is there anythin like what i'd be looking for?

[Alastor187]

Originally Posted by pik_d

i've had the crazy idea of writing a (c++) program which would make a pretty display of HP and torque based on some form of input.

the most obvious way i thought of was to input either the hp or torque at certain intervals along the powercurve. but i dont really like that idea, to much to input to make it look smooth.

my next idea was to find some calculation which would take "SOMETHING" and do some nifty calculations and give power/torque numbers. yay.

now, for all you guru's, is there a way to take bore, stroke, cylinder numbers, compression ratio, etc, and make a ROUGH guess of horsepower and torque?

i looked around a bit, and found something called RAChp (royal automotive club, from england, also called taxable horsepower) which is this equation:

D^2 * (n*2.5)

where D is the bore in inches, and n is the number of cylinders.

now, this is obviously NOT going to be close at all, but it's all i've found. (by that, an SRT-4 gets 136hp. )

so... is there anythin like what i'd be looking for?

I think the following set of equations should get you started:



Summary –

EngineDisplacment * EngineSpeed = IdealAirFlowRate

IdealAirFlowRate * VolumetricEfficiency = ActualAirFlowRate

ActualFlowRate * AirDensity = ActualMassAirFlowRate [I suggest Using A Standard Density]

ActualMassAirFlowRate / AirFuelRatio = ActualMassFuelFlowRate

LowerHeatingValue * ActualMassFuelFlowRate = FuelEnergyRate [Assume Water Vapor]

1 – ( 1 / CompresionRatio^0.4 ) = ThermalEfficiency

ThermalEfficiency * FuelEnergyRate = PowerOutput



If you choose constants for VE, TE, etc the resultant power output will be linear with respect to speed.

Obviously, there are many other parameters that affect performance most of which can be ignored but you probably want to develop a few simple equations for heat loss and friction.

Heat loss will be inversely proportional to the square of the engine speed, while fiction will be directly proportional to the square or cube of the engine speed. In other words:

PowerOutput – [HeatLoss / ( EngineSpeed^2)] – Friction * EngineSpeed^2 = NetPowerOut

[curtis73]

Not that I suggest theft but you might get a good idea of where you're going by looking at proracingsim.com . They have the Desktop Dyno line of simulation software that is exactly up the alley you're thinking.

[pik_d]

thanks for that guys... i'll read over it and attempt to make heads and tails of it tomarrow... 5am now and nothin doin for me.

[pik_d]

Originally Posted by Alastor187

If you choose constants for VE, TE, etc the resultant power output will be linear with respect to speed.

Obviously, there are many other parameters that affect performance most of which can be ignored but you probably want to develop a few simple equations for heat loss and friction.

Heat loss will be inversely proportional to the square of the engine speed, while fiction will be directly proportional to the square or cube of the engine speed. In other words:

PowerOutput – [HeatLoss / ( EngineSpeed^2)] – Friction * EngineSpeed^2 = NetPowerOut

hmm, sorry to double post, but it's been half a day, and i've got stuff to add.

i probably will just use constants for VE and TE, and of so, what number should i use?

so, as far as actually finding numbers for heatloss and friction, how should i do that?

[Alastor187]

Originally Posted by pik_d

hmm, sorry to double post, but it's been half a day, and i've got stuff to add.

i probably will just use constants for VE and TE, and of so, what number should i use?

so, as far as actually finding numbers for heatloss and friction, how should i do that?

For conservative, but not overly so, values of VE and TE I would use 80% and 30%, respectively.

As far as has the heat and friction losses I don’t have any suggestions at the moment. I will look into it and try to find some realistic values that scale properly. Hopefully, in the meantime someone else will be able to suggest sufficient values.

For clarification I suggested including the heat and friction losses so that the power curve would not be linear with respect to RPM, and therefore ‘look’ more realistic. I realize you are only trying to make a very simplified model, but a straight power curve would just be silly.

It may be easier for you to ignore the heat and friction losses and instead vary volumetric efficiency. Instead of inputting a constant make it dependent on RPM, Google should be able to find a suitable curve to model after:

http://images.google.com/images?hl=e...cy&sa=N&tab=wi

The only issue I see with this is that the resultant power curve will follow the VE curve. Generally, it is the torque curve that follows the same trends as the VE curve. Nevertheless, if this sounds like a more reasonable route let me know and I think we can modify the original model to make this work.

[pik_d]

i have a few more questions

air density... is that measuerd in kg/m^3, psi, what? and what standard number should i use?

this site claims that the LHV of gas is 43 MJ/kg, and the air fuel ratio is 14.6. are these values that i can use as constants?

[pik_d]

ok. so i kinda got it to work. except the power output values are in the hundred thousands...

i'm pretty sure it's just that i'm using the wrong units. i've got all the math to what those image files require (with all the constants), but i just dont know what units to use...

[Alastor187]

Originally Posted by pik_d

i have a few more questions

air density... is that measuerd in kg/m^3, psi, what? and what standard number should i use?

this site claims that the LHV of gas is 43 MJ/kg, and the air fuel ratio is 14.6. are these values that i can use as constants?

I believe the standard air density at sea level is about 1.229 kg/m^3.

That constant LHV of 43 MJ/kg should be sufficient. Generally, the A/F ratio is higher at low loads and lower at high loads. The exact values will vary between different engines, an example fuel map can be found here:

http://www.stealth316.com/images/fue...5spydervr4.gif

Using a constant of varying value for the A/F ratio depends on how complex you are willing to make your model. If you want to stick with a constant value 14.7 would be a reasonable choice.

[Alastor187]

Originally Posted by pik_d

ok. so i kinda got it to work. except the power output values are in the hundred thousands...

i'm pretty sure it's just that i'm using the wrong units. i've got all the math to what those image files require (with all the constants), but i just dont know what units to use...

Using the correct units is definitely important.

Can you post your inputted values and corresponding units?

[pik_d]

EngineDisplacment * EngineSpeed = IdealAirFlowRate

i input bore and stroke (in inches), and it calculates out the engine displacement correctly. engine speed... i just have it going every 500rpm's in an array. idealAirFlowRate is the just the raw number produced by the multiplication of these two numbers (cubic inches x rpm)

IdealAirFlowRate * VolumetricEfficiency = ActualAirFlowRate

VE is set to .8 and multiplied by the afore mentioned raw number to get yet another raw number

ActualFlowRate * AirDensity = ActualMassAirFlowRate [I suggest Using A Standard Density]

multiplying the last number i got by 1.225kg/m^3

ActualMassAirFlowRate / AirFuelRatio = ActualMassFuelFlowRate

dividing last number by 14.7

LowerHeatingValue * ActualMassFuelFlowRate = FuelEnergyRate [Assume Water Vapor]

multiplying last number by 43. (not sure if this should be a higher or lower power)

1 – ( 1 / CompresionRatio^0.4 ) = ThermalEfficiency

inputting compression ratio (for example, 9.8:1 ratio is inputted as simply 9.8) and finding TE with that

ThermalEfficiency * FuelEnergyRate = PowerOutput

multiplying the two numbers to get a power output.

i input:
bore: 3.44
stroke: 3.27
cylinders: 4
engine redlin: 6500
compression ratio: 9.8

output:
500rpms yields a number of 105364, for example...
5000rpms yields 1053718

[Alastor187]

First order of business is to choose system of units (DO NOT MIX UNITS), the two most common are US customary (also referred to as Imperial) and SI. The standard follows:

For USC Units:

Distance = Feet [ft] or Inches [in] (depending on the calculation, often a source of error)
Mass = slugs [slug] or pound-mass [lbm] (again, often a source of confusion)
Force = pound-force [lbf]
Time = seconds [s]
Energy = British Thermal Unit [BTU]
Work = feet-pound [ft-lb]
Torque = pound-feet [lb-ft] (notice it is not the same as work)
Power = horsepower [hp]

For SI Units:

Distance = meters [m]
Mass = kilograms [kg]
Force = Netwons [N]
Time = seconds [s]
Energy = Joules [J]
Work = Joules [J] ??? ( I am not 100% sure if this correct)
Torque = Netwon-meter [Nm]
Power = Watts [W]

Those are some of the most basic units, and should be enough for this analysis. I would suggest using the US customary units if you are looking for horsepower as an output. If you are having trouble converting between units, try using the Google calculator. All you have to do is go to www.google.com and type "X unit1 to unit2". For example a conversion of 1 meter [m] to feet [ft]:

http://www.google.com/search?hl=en&l...ft&btnG=Search

Originally Posted by pik_d

i input bore and stroke (in inches), and it calculates out the engine displacement correctly. engine speed... i just have it going every 500rpm's in an array. idealAirFlowRate is the just the raw number produced by the multiplication of these two numbers (cubic inches x rpm)

It is fine if your bore and stroke inputs are in inches, but convert them to feet in the code. Working in feet, the bore area times the storke will give you a volume in this case cubic feet (ft^3). Multiplying your cylinder volume times the number of cylinders gives your total engine volume in cubic feet.

BoreArea * Stroke = CylinderDisplacement => ft^2 * ft = ft^3
CylinderDisplacement * NumberOfCylinders = TotalDisplacement => ft^3 * 1 = ft^3

RPM can be broke down in to “revolutions / minute”, where revolutions can be considered unitless. Therefore, if you multiply engine speed by engine displacement the results will be cubic feet per minute.

RPM = Revolutions / Minutes = 1 / Minutes
RPM * TotalDisplacement = IdealAirFlowRate => 1 / Minutes * ft^3 = ft^3 / minute

Since the standard unit so time is seconds, you need to convert the flow rate from cubic feet per minute into cubic feet per second (multiply by 1 / 60).

===

If you are still having problems I can go through the calculations and corresponding units in more depth. But for now it is off to bed.

[curtis73]

I love you guys. These are intelligent and informed posts... and way beyond what I know

I'm glued to the computer screen following all the math.. or at least trying.

[pik_d]

curtis, some of this stuff is over my head... so i just look at it as numbers and equations i just need to figure out how to program in.

alastor-
ok. my numbers are alot better... except i have a neon engine putting out over 300hp at 6000rpm's

EngineDisplacment * EngineSpeed = IdealAirFlowRate

i input bore and stroke (in inches), divide by 12 to get feet, and have it calculate the engine size in cubic feet. it does so correctly.

engine speed is same as last time, except divided by 60 in the equation to find ideal airflow rate

IdealAirFlowRate * VolumetricEfficiency = ActualAirFlowRate

VE is set to .8 and multiplied by the afore mentioned raw number to get yet another raw number (unchanged)

ActualFlowRate * AirDensity = ActualMassAirFlowRate [I suggest Using A Standard Density]

multiplying the last number i got by .0751265 lbf/ft^3

ActualMassAirFlowRate / AirFuelRatio = ActualMassFuelFlowRate

dividing last number by 14.7 (unchanged)

LowerHeatingValue * ActualMassFuelFlowRate = FuelEnergyRate [Assume Water Vapor]

multiplying last number by 18486 btu/lb (i converted 43MJ/kg, not sure if this is the right number to use)

1 – ( 1 / CompresionRatio^0.4 ) = ThermalEfficiency

inputting compression ratio (for example, 9.8:1 ratio is inputted as simply 9.8) and finding TE with that (unchanged)

ThermalEfficiency * FuelEnergyRate = PowerOutput

multiplying the two numbers to get a power output.

i input:
bore: 3.44
stroke: 3.27
cylinders: 4
engine redlin: 6500
compression ratio: 9.8

output:
500rpms yields a number of 25.46
5000rpms yields 264.2

[Alastor187]

I put together a quick spread sheet using your numbers, and got slightly different power outputs. It is possible that I made a mistake so please double check my numbers against yours.

First, I contradicted myself earlier in regards to the thermal efficiency. I told you to use 30% while the same time giving you an equation to calculate TE from the compression ratio. Well, the equation I gave, although mathematically correct, will give you too high an efficiency about 60% when in the real world it will be about half of that.

With that said lets look at the results.

I am using the following dyno sheet for reference:

http://www.zex.com/Base/Images/DynoSheets/107-200-8.gif

My spread sheet results are:



Columns A – C are for 3400 rpm, while columns E – G are for 5800 rpms. Columns B and F are the results with a thermal efficiency of 60% (calculated from the compression ratio). Columns C and G are the results using a thermal efficiency of 30% and give power outputs very close (less than 10% difference) to the dyno values.

If my numbers are correct I would say this simple model is fairly accurate, more so than I would have imaged. This is probably because all the simplification that have been made balance each other out.

For example, we have neglected fiction losses that lower the engines net power output. While at the same time during high engine loads the A/F ratio will decrease causing an increase in the engines net power output. If these two characteristic balance each other equally then the results can be relatively accurate. However, there is no way to know for sure with out extensive testing, it is just a matter of luck. As they say “it is better to be lucky than good”.