Machin's Performance prediction calculations

Post here all non technical related topics about Formula One. This includes race results, discussions, testing analysis etc. TV coverage and other personal questions should be in Off topic chat.
User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Here's the same car model with an engine that can make 600bhp over 2000rpm (between 10,000rpm and 12,000 rpm), the output (green) now fills in the area highlighted red by 747 above....

Image


And the acceleration now looks like this (compare to the very first graph on page 1 of this topic).

Image

What we want to do now is lose the "flat spots" in the acceleration curve where the driver is changing gear.... (FYI I've optimised the gear ratios which is why the flat spots occur at different speeds than they do in the graph on page 1).
Last edited by machin on 21 Dec 2010, 14:55, edited 1 time in total.
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

User avatar
747heavy
24
Joined: 06 Jul 2010, 21:45

Re: Machin's Performance prediction calculations

Post

Yes, I agree
Ideally you would operate your engine (depending of the application) either at the max power point or at the max efficiency point, and let the CVT do the rest.
If you can tune your engine, to give a range of both, you gain some freedom, as a belt CVT does not have infinite ratio in the real world, and the efficiency depends (among other things) from the offset/ angle of the belt.
But, yes there is a cross over point where you regain efficiency fom the engine and offset it against the lower efficiency of the CVT.

For this reason I would like to see F1 allow gastubines and CVT again.
I´m sure it would be a doable alternative to the turbo I4.
Let´s both have the same amount of fuel, and the better concept should win - IMHO
"Make the suspension adjustable and they will adjust it wrong ......
look what they can do to a carburetor in just a few moments of stupidity with a screwdriver."
- Colin Chapman

“Simplicity is the ultimate sophistication.” - Leonardo da Vinci

User avatar
747heavy
24
Joined: 06 Jul 2010, 21:45

Re: Machin's Performance prediction calculations

Post

good job machin !!!

I think we come now closer to what I would expect to see in 2013.
I´m not enough of an engine guy to say if 10000-12000 would be the optimum
range, all things considered friction losses, inertia and boost levels -> charge temp etc.

I would perhaps expect to see it a little bit lower in the rpm range.
Maybe 7000-9000 rpm to minimize friction losses and inertia, but this is not really my strong suit, so I will leave it to the engine guy´s.

But the principle holds true - IMHO.
This engine/gearbox combination, would not benefit much from CVT - IMHO
What is the shift time you use in you model Machin?
Maybe we can cut down on this a bit with things like "ZeroShift" etc.
Could you overlay the last accle graph with a graph for "constant power"(ideal) and one for the constant torque engine - please?
(Only if it is not too much hassle, no need to spend hours on a new model etc.)
"Make the suspension adjustable and they will adjust it wrong ......
look what they can do to a carburetor in just a few moments of stupidity with a screwdriver."
- Colin Chapman

“Simplicity is the ultimate sophistication.” - Leonardo da Vinci

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Maybe we can cut down on this a bit with things like "ZeroShift" etc.
Yeah -I use 0.1seconds in this model......

No worries:- Here is the difference between a constant power engine with 1 gear and the constant torque engine with gears (the other option would be between the two). This assume that the cars are exactly the same other than the engine output... so it assumes that they have the same rotating inertia etc.

Note that since the constant torque engine must have a really low gear for low speed acceleration the effect of the engine/clutch/gearbox input inertia is higher (it must accelerate through more RPM for a given increase in road speed) which is why the constant power engined car accelerates slightly quicker at really low speeds (0-60mph), despite having the same grip available.....

Image
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

autogyro
autogyro
53
Joined: 04 Oct 2009, 15:03

Re: Machin's Performance prediction calculations

Post

Would the 25 gear ratio bevel epicyclic fully automatic clutchflite transmission I was working up in 1977 be any use?

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Auto, Possibly, but this chart shows you only need 7 gears and an engine capable of constant power over 2000rpm to pretty much maximise acceleration... any more gears and you're just wasting time changing between those gears....

Image
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

autogyro
autogyro
53
Joined: 04 Oct 2009, 15:03

Re: Machin's Performance prediction calculations

Post

machin wrote:Auto, Possibly, but this chart shows you only need 7 gears and an engine capable of constant power over 2000rpm to pretty much maximise acceleration... any more gears and you're just wasting time changing between those gears....

Image
Yes and that is why I suggested the 7 gear maximum to Jordan and it became the regulation maximum.
However, if you design the engine for a narrower but higher power band the number of gears increases.
You do of course need a much faster shift than a layshaft box, even with trick shifts. (which are in effect no faster than ordinary dog clutch changes, just more reliable hopefully).
All comes round to a new generation of geartrain that takes into account hybrid and full electric application.
I knew this in 1977.

User avatar
ringo
230
Joined: 29 Mar 2009, 10:57

Re: Machin's Performance prediction calculations

Post

The issue with the single gear. Is that theoretically it's makes no difference, with multiple gears, but when the inertia of the car and parts, and external forces are considered it's best to use more than 1 gear.
We already made it clear that yes, with constant power, 1 gear will be enough to power the car up to it's terminal speed. But it only equals a CVT or stepped gearbox if and only if, we ignore the resistance to motion of the car. The acceleration is the elephant in the room.
The acceleration of the car is a combination of things. Also the acceleration is not constant becuase that is only the case with constant torque. In our case torque is constantly changing in a constant power car.
Find the moment of inertia of the rotating gears, shafts, differential and wheels. Then the linear component of acceleration can be considered as well, in terms of the car itself and the other external forces. I don't have time to do all this now, but just putting it out there. I'll try it soon.
The difference between the drag, resistance etc. should be total external forces, even the weight on an uphill if needs be. That can be used against the wheel forces to find the acceleration of the wheels and car.

To actually calculate the time to accelerate, or how fast to red line, plot the calculated angular acceleration against the engine angular velocity, and find the slope of the curve at different points to get the time segments. Add the segments to get the total time to red line in each gear.
This way a drag race can be simulated.
For Sure!!

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Seriously dude: read my first post on this thread again; my acceleration calculation does take into account inertia of the wheels, tyres, brakes, drive shafts, diff gears, transmission gears and the engine. I know what I'm talking about!

The inertia is actually worse in lower gears as u must accelerate the engine components much faster for a given change in road speed... So actually add more lower gears to a truly constant power engine and acceleration will be worse! Only a small amount, but a small amount none the less.

Mates yet?!
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

It also has a component for gradients (positive and negative), but i've assumed a flat road for all these graphs.
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

User avatar
ringo
230
Joined: 29 Mar 2009, 10:57

Re: Machin's Performance prediction calculations

Post

How do you find your acceleration?
I would like to see a sample calculation, out of curiosity.
For Sure!!

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

The way I do the inertia is to equate the inertia of the engine and transmission to an equivalent inertia at the road wheels...by taking into account the relative rotational speed of there components compared to the road wheels... U need to work out the inertia separately for each gear therefore since the relative speeds are different... My program does that all for u, and I've created my own system for estimating the inertias of there parts based on a load of different factors and a survey of many real parts (I'm by geek, what can I say!)... Or I can individually add in actual values if they're known...

Convinced I know what I'm talking about now?!
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Honestly, give it a go yourself, what you'll find as I did is that the effect of the rotating parts is fairly small compared to the mass of the car... And in addition its worse in low gears due to the higher ratios...
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH

User avatar
ringo
230
Joined: 29 Mar 2009, 10:57

Re: Machin's Performance prediction calculations

Post

machin wrote:The way I do the inertia is to equate the inertia of the engine and transmission to an equivalent inertia at the road wheels...by taking into account the relative rotational speed of there components compared to the road wheels... U need to work out the inertia separately for each gear therefore since the relative speeds are different... My program does that all for u, and I've created my own system for estimating the inertias of there parts based on a load of different factors and a survey of many real parts (I'm by geek, what can I say!)... Or I can individually add in actual values if they're known...

Convinced I know what I'm talking about now?!
Ok, just checking to see if you really know what you are doing. But the best direction to go initially is to find the net force on the car first, subtract the drag and rolling resistance from the wheel force at that speed (assuming C.O.P is at wheel centre height), to get the net torque on the wheels. It wont change the relations, just the acceleration.
I am doing it now. Now remember as i said, there is something that is not being accounted for, i think it could be the initial slip. I have already taken note of the single gear ratio with constant power, and yes i can see where you said 1 gear works. :roll: The only issue is the first movement from idle. :wink:
This is based on an idle speed.
However both multi and single gear solutions are the same if both start at an engine speed of zero. Otherwise the multi gear has an advantage.

The reason, both are equal from zero, is that coming from an infinite torque nothing can really prevent one from not moving. Only in the case of starting from a cut off point with an initial resistance, will there be a need for a threshold force to start motion.
In the case of an arbitrary engine idle speed, the lower gear could mean the difference between moving off and stalling. Maybe on a slope i suppose. Rolling resit, isn't so great to make that difference.
And as i mentioned the slip already, the higher gear has a bigger wheel speed and and road speed differential as the clutch is let in compared to the lower gear, given an idle speed.
The wheel will slip for an amount of time; this artificially gives the multiple geared car a head start.
In my data my first gear gives 7m/s at 1000 rpm, my 7th gives 38m/s.
Going off throttle would aleviate this as you say, but off throttle defeats the purpose of constant power.
This is also why i realize constant power doesn't make sense.
For Sure!!

User avatar
machin
162
Joined: 25 Nov 2008, 14:45

Re: Machin's Performance prediction calculations

Post

Seriously though Ringo... Please read my very first post again... I already do exactly what u say... I calculate the net force at the driven wheels, and its this net force that I use with the mass and the inertia to calculate the acceleration... 'simples'.

Going 'off throttle' is normal... In the old 3.0 litre, 20,000rpm+ days they couldn't use full throttle Til over 100mph when enough grip was generated from the downforce!
COMPETITION CAR ENGINEERING -Home of VIRTUAL STOPWATCH