How they should measure hp and torque gains from mods.
- Thread starter sleeper_
- Start date
7
70
I'm not really cathcing what you are saying here. Can you maybe explain it a different way?
I don't think thats such a good idea. It's kinda hard for most people to quatify areas under curves, espcially if they don't know what the curve looks like orginially. It works great for people that do dyno testing all day every day cause they understand the principles behind it, most other people it's just easier to state it in terms of HP and Ft/Lbs gained/lossed.
PA_MP3_Man
Member
Little off topic but StuttersC, Is the a Dawsons Creek Futura S2000 Trapper Keeper? 
well.. it wouldn't be that difficult to calculate area (i'm sure computers can be programmed to do it quite easily).. if a 2 dyno graphs are taken.
also, of course the normal dyno graphs should also be given to keep it within an optimum power rpm range, but for STRAIGHT line speed, area above the original stock dyno graph should have subtracted from it the area below it. since people would be using the full rpm range for every gear in an acceleration that is.
i'll explain later maybe, but now i gotta jet. lol laters
also, of course the normal dyno graphs should also be given to keep it within an optimum power rpm range, but for STRAIGHT line speed, area above the original stock dyno graph should have subtracted from it the area below it. since people would be using the full rpm range for every gear in an acceleration that is.
i'll explain later maybe, but now i gotta jet. lol laters
7
70
PA_MP3_Man said:Little off topic but StuttersC, Is the a Dawsons Creek Futura S2000 Trapper Keeper?
Yes, it is the Dawsons Creek Trapper Keeper Ultrakeeper Futura S2000...
That epsiode is classic..."Trapper Keeper, hybrid with Wellington bear Calculator."
PA_MP3_Man
Member
That was a great episode. I love south park. Its like one of the best shows ever. New episodes start nov. 6. Can't wait man.
kooldino, that of course is useful to know where in the rpm range the power gain lies, HOWEVER, when comparing two mods with comparable performance gains, one cannot clearly see which mod or product produces more power (for straight line speed). Area under the curve can be calculated by interpolating values between "known" points or plotted points of the graph. If you've taken calculus you should know that there are different methods of interpolation, the simplest one using the midpoint of the given x range and y range and multiplying it by the y value to estimate the area for the specified x interval.
to put what i'm proposing into practice think of this: two cai's from different manufacturers where brand A claims peak power gain of 5whp and brand B claims peak power gain of 7wph, but upon looking at the graphs of the two products you can't really tell which one is actually "better" while it is easy to assume brand B of 7whp gain is better when in fact it may not be. Brand A may in fact have a new torque curve that the total area between it and the stock torque curve totals to a greater # than that of brand B's. This translates into a faster straightline times with brand A vs brand B. HOWEVER, in an autocross type situation where you're not using the whole rpm range you might find that product B of 7whp peak gain is in fact more advantageous b/c you would be aiming to keep the engine in it's peak powerband for the duration of the autocross. But then what i propose is again the same for comparing products. To compare the areas of the segment of the rpm range you're using most at any time allows you to clearly determine which product is superior (ie, brand A still might be better than brand B in this case).
to put what i'm proposing into practice think of this: two cai's from different manufacturers where brand A claims peak power gain of 5whp and brand B claims peak power gain of 7wph, but upon looking at the graphs of the two products you can't really tell which one is actually "better" while it is easy to assume brand B of 7whp gain is better when in fact it may not be. Brand A may in fact have a new torque curve that the total area between it and the stock torque curve totals to a greater # than that of brand B's. This translates into a faster straightline times with brand A vs brand B. HOWEVER, in an autocross type situation where you're not using the whole rpm range you might find that product B of 7whp peak gain is in fact more advantageous b/c you would be aiming to keep the engine in it's peak powerband for the duration of the autocross. But then what i propose is again the same for comparing products. To compare the areas of the segment of the rpm range you're using most at any time allows you to clearly determine which product is superior (ie, brand A still might be better than brand B in this case).
7
70
Ok, I see what you are getting at now...
I know finding the area under a curve is bsically calculus...And you think that instead of comparing simply numbers and or just the curve itself, you want to compare the area under the curve because as you said, product A might actually have more "area" under the curve than porduct B. And that having more "area" under the curve might prove to be a better product because it increases power throughout the power band instead of at just a peak specific point.
Have fun doing the calc, I'll just guesstimate with my little eye as to what might be better given the dyno graphs supplied by various venders.
I know finding the area under a curve is bsically calculus...And you think that instead of comparing simply numbers and or just the curve itself, you want to compare the area under the curve because as you said, product A might actually have more "area" under the curve than porduct B. And that having more "area" under the curve might prove to be a better product because it increases power throughout the power band instead of at just a peak specific point.
Have fun doing the calc, I'll just guesstimate with my little eye as to what might be better given the dyno graphs supplied by various venders.
You cant calculate the area under a curve unless you are given an exuation for that curve, or if you have some type of calculus program. Mathlab might do it. But calculating the area under a curve is easy, all it is is an integral. But the hp curve doesn't really have an equation to go with it. You don't get dyno chart with something saying, y=2x+8x^3/89
lol.. nonono i know with calculus and areas you're basically finding the sum of the infinitely thin slices under a curve. HOWEVER there is area approximation as i explained above. you dont' NEED an equation of the torque curve, it's just simple number crunching: you divide the curve up into intervals (depends on sample rate) and calculate approx area that each interval gives and sum the areas up.
eg. coordinates (3,3) and (5,5):
x midpoint = 4
y midpoint = 4
midpoint --> (4,4)
approximate area of the x interval 3 --> 5 is 4x2=8
so 8 cubic units becomes the approximate area of the interval, the curve, and the x-axis.
do this for every interval (depends on the sample rate of the dyno computer if it takes a reading every 1rpm or 10 rpms etc) and sum them all together and you get an "approximate" area that's relatively accurate. Accuracy is limited by the sample rate of the ecu/dyno computer; HOWEVER, considering the error factor or "variability" of the equipment and conditions etc i doubt a little change in the sample rate (every 10rpms vs every 5) would make a significant contribution to the accuracy of the area. Every 1000 rpms vs 100 rpms in fact would make a difference but of course that's not the case.
eg. coordinates (3,3) and (5,5):
x midpoint = 4
y midpoint = 4
midpoint --> (4,4)
approximate area of the x interval 3 --> 5 is 4x2=8
so 8 cubic units becomes the approximate area of the interval, the curve, and the x-axis.
do this for every interval (depends on the sample rate of the dyno computer if it takes a reading every 1rpm or 10 rpms etc) and sum them all together and you get an "approximate" area that's relatively accurate. Accuracy is limited by the sample rate of the ecu/dyno computer; HOWEVER, considering the error factor or "variability" of the equipment and conditions etc i doubt a little change in the sample rate (every 10rpms vs every 5) would make a significant contribution to the accuracy of the area. Every 1000 rpms vs 100 rpms in fact would make a difference but of course that's not the case.
oh, i know, you weren't attacking you just simply misunderstood..
i dont' know if dynos do that for you, BUT i bet they could be programmed to do that w/o that much work.
as for the "wtf" i was meant as "wtf, i never would do all that work and i'd be crazy to suggest it" type sorta thing. hahaha...
i dont' know if dynos do that for you, BUT i bet they could be programmed to do that w/o that much work.
as for the "wtf" i was meant as "wtf, i never would do all that work and i'd be crazy to suggest it" type sorta thing. hahaha...
I don't really know what programs dyno's use either. They may have an area that regular mechanics don't know about that will do the integral for you. My best guess is that most mechanics don't know what calculus is and they wouldn't know how to do it. An easy way also is to import the curve into something else like microstation or autocad and just flood tha area and measure it that way.
