General Chat
Q:
So we all know when we talk about the HP of a car we are talking about the peak HP. Then people often take the peak HP and divide it by the wieght of the car to give a hp per pound figure. Well that only shows you which car has more power at the peak, which isnt nessesarily even at the same rpm.
So really whenever you compare the peak hp numbers its really just comparing apples and oranges.
You tell me wich car is faster? A car that makes more HP early in the powerband than another obviously has an advantage. This is where the idea of the dyno chart comes in, we need a graph of the powerband to tell which car has more power, and therefore is faster. So lets imagine two dynos, both are of the same car, with the same wieght, same tires, etc, the only difference is the powerband they are making. Lets look at what this dyno could look like.
Now do you see how the graph crosses? The red graph winds up being greater, so that car would have higher peak HP, but not nessesarily be more powerfull overall.
Which leads us to the idea of the area under the curve(graph). The area shaded by red shows when the black dyno is greater than the red. And the yellow when the red dyno is greater than the black.
When you compare the shaded area of the red and yellow you can see that the red area is slighty greater than the yellow area. This means that the car with the black dyno graph has greater power overall than the car with the red dyno graph, despite the fact that the red dyno graph has a higher peak HP.
Now thats all fine and dandy, thats a nice simplified case not likely to occur in the real world. So this leads us to the dyno information. From the data file of a dyno we can export the data points from the dyno to a text file, MS Excel can take a text file with that data and find a graph and a regression for that data. This will give us the equation for the dyno.
Here is an example of 2 cars compared this way:
As you can see the plot of the data for a dyno vs the equation of the line isnt a perfect match. If we took a higher order polynomial we could find a more accurate equation. But for this case this is accurate enough. The r^2 value correlates to the accuracy of the graph. The closer the value to 1 the more accurate the graph.
Now from these 2 equations, y = -2.1274x6 + 59.635x5 - 679.12x4 + 4003.1x3 - 12826x2 + 21237x - 14158, and y = -3.1937x6 + 91.167x5 - 1061.2x4 + 6425.1x3 - 21280x2 + 36611x - 25533. We can use calculus and the ideas of integration to find the total area under the curve. However we are presented a problem, to do this we must compare the area under the curve for both graphs over the same powerband. This means that we need to take the area from 2.5krpms to 6.5krpms, or whatever values we want, as long as we have data for those values.
So by doing this we find the total HP for the first dyno graph is 1561.63HP. And the second is 1404.89HP(both between 2.7 and 6.9krpms) Thats a total difference of 156.74HP. So since both cars are the exact same except for the HP the first car will be faster.
Now if both cars werent the same we could simply divide the actual wieght of each car by its total horsepower and find the HP/wieght ratio, whichever car had the better ratio would represent the faster car(in ideal conditions, not taking into account drivers, tires, shifting....etc etc)
This by far gives us the best information about what car is faster and has more HP.
Currently own:
1990 2+2(For sale soon)
1991 TT(parts for sale)
1993 NA(parts for sale)
1993 Convertible(Converted TT)
1996 TT(with 9,124 miles on it)
"When you do things right people cant really tell that you've done anything at all."
l
A:
All that thinking and all that math proves nothing on the track where the rubber meets the road.
Sure HP numbers are fun to quote for bench racing, but until you have traction and a driver, all you have is numbers.
Excellent points though for conversation.
www.zselect.com
75 280Z Current project car
75 280Z Hybrid Chevy 350
90 300ZX Just for fun
A:
too much math....ugh
-----------------------------------------
12.804 @ 107.26 mph on crappy street tires with Dual SU Powered 2.9L Stroker!
ZCAR.COM member since Aug 1998
A:
The point is to level the playing field.
sure two brand new cars that are the exact same being raced against each other will come down to a drivers race. This is meant to level the playing field...
too often someone will race and say...well I missed a shift, or I couldnt hook up the traction, and those are used as excuses for losing.
well by taking the dyno we can elminate the driver and say...
"look if your did everything 99% perfect heres the outcome....."
THEN on top of that you can consider a more experienced driver, or someones mistakes while driving.
Currently own:
1990 2+2(For sale soon)
1991 TT(parts for sale)
1993 NA(parts for sale)
1993 Convertible(Converted TT)
1996 TT(with 9,124 miles on it)
"When you do things right people cant really tell that you've done anything at all."
l
A:
Good point 280zone considering I watched 01 cobra run a 15.30 over and over the other day.
-1986 z na to turbo conversion
-JWT 550 fuel kit
-JWT sport 500 ball bearing
-NA block
A:
I aint no fancy college boy.. wait - yes I am. No engine in the world has a hp graph that rises then lowers than raises. That is impossible.
A:
my brain hurts :p
___________________________________________________________
Erik Coleman
Orange '76 280Z 2+2
Red '85 300ZX Turbo
Black '86 300ZX NA
A:
I'd have to agree that you probably won't see a graph like that. But, the concept is the same on a real graph.
The guy that makes the most power for the longest period of time would have an advantage, even if he has less peak Hp. But that still doesn't mean squat until the rubber meets the road as to who will win in a drag race, or a lapping contest.
You have to analyze the WHOLE dyno chart, not just the last little bit.
Which brings me to something I don't ever see mentioned here, and that's timing curves. Everybody always talks about static and peak advance but no one seems to be paying attention to 'how' it got to max advance. Did it come in fast? Slow? What did the curve look like?
Use to, back in the day, everyone knew where a guy was that had a distributer machine.
Ken
'82ZX n/a 2+2
'02 Sportster XL1200C
A:
that you would ever see a graph that looked like that, but its the concept that im showing by taking it to extremes.
Currently own:
1990 2+2(For sale soon)
1991 TT(parts for sale)
1993 NA(parts for sale)
1993 Convertible(Converted TT)
1996 TT(with 9,124 miles on it)
"When you do things right people cant really tell that you've done anything at all."
l
A:
...ummmm....and from all this I can derive that I am fully and completely lost.
______________________________________________________________________________
77 280Z - TEC3 Engine Management, Coil Packs, 3.1 Stroker, 550CC injectors, tripple throttle body FI intake, MSA Stage II Cam, etc...etc...
79 Fairlady - Rusty
81 280ZX - DD
A:
Where is your differential integration?
Don't make me bust out some calculus on you.
but basically you need to use the Reiman Sum from a given interval...we'll say 2.5 to 7.
You can either just do normal rectangle integration, trapeziod integration, or my favorite, Simpson's rule...
Basically you break that graph up in .5 width rectangles...
Then you take the sum of the limit of those integrals as N approaches Infinity...
so f(xsub0)(DeltaX)+4(f(xsub1)(DeltaX)+2(f(xsub2)(DeltaX)....etc etc....
That will get you the area of the graph....do that for both, and subract the smaller one from the larger one to find the area difference.
...
Post Edited (Nov 17, 11:44am)
A:
I did the integration. Those numbers given are the integrals for the regression calculated with excel....(from 2.7 to 6.9k rpms)
No reason to do a reimann sum, thats the whole point of getting the equation. Look back, I think you just missed where that step occured. I skipped over the actual integration since this isnt a math lesson but is a concept lesson(the math is secondary)
Currently own:
1990 2+2(For sale soon)
1991 TT(parts for sale)
1993 NA(parts for sale)
1993 Convertible(Converted TT)
1996 TT(with 9,124 miles on it)
"When you do things right people cant really tell that you've done anything at all."
l
A:
I think most people have the advance in all the way by the time the engine is making useable power on these cars, which is around 2000-2500 even with a mildly tuned version.
Steve 77 280z, 3.1L, SDS w/turbo injectors, MSD, 280/.510cam, titanium retainers, Sunbelt-intake and head, Nismo1.75 headers, susp mods, toyo calipers etc etc.. my engine
A:
The integration is easy...its that nasty polynomial thats hard to find! Even with Simpson's rule you still need to find the polynomial. I'm guessing some sort of computer program was used to find it. I think the easiest thing would be to find multiple points on the curve, connect the dots, and use a trapezoid or midpoint estimation of the area.
You know what would be even better...instead of having HP as a function of RPM:
HP(RPM) = A(RPM)^(n) + B(RPM)^(n-1) + C(RPM)^(n-2)...
...Have HP as a function of time:
HP(T) = A(T)^(n) + B(T)^(n-1) + C(T)^(n-2)...
Then the integral of HP(T) would be the kinetic energy (K) of the car:
K(T) = (A/n+1)(T)^(n+1) + (B/n)(T)^(n) + (C/n-1)(T)^(n-1)...
Via the kinetic energy equation, and knowing the mass (M) of the car:
K=(1/2)(M)(V)^2...
The velocity (V) of the car as a function of time (T) would be:
V(T) = Sqr Rt(2K(T)/M)
Substituting K(T) into V(T) would be a pain in the arse...so I'll leave that to those who are interested enough to do it themselves.
To take shifting gears into account, you would have to break this up into multiple integrals with changing bounds of integration.
Aaaaah!!! Too much work! Just get a stopwatch and clock your best 0-60 MPH and 1/4 mile and be done with it!
1971 240Z
HLS3026814
A:
give it a break. you have proved you points.
----------------------------------------------------
1981 n/a 280zx
1952 m38a1 jeep willy's
1971 fairlady 240z
95 300zx N/A
72 240z
A:
What about gearing?
-------------------------------------------------------------------
"No it's not for sale, I'm going to be fixing it up soon."