Honda D Series Forum banner
1 - 20 of 142 Posts

· Registered
Joined
·
4,359 Posts
Discussion Starter · #1 ·
After hearing/reading so much misinformation regarding torque and horsepower, I figured this write-up was needed. Although I have to go over some basic physics for this article, I’ll keep calculus out of it. Let’s start by going over the relevant terms:

Thrust
Thrust is the force that drives your car forward. It doesn’t matter how much torque your engine makes, what your gearing is, how large your wheels are, how shiny your muffler is, or how many stickers you have on your doors – thrust determines how fast your car can accelerate. In racing, acceleration is what really matters, and there are only a few immediate ways to make your car accelerate faster:

ΣF = ma
Where, ΣF= the sum of the forces acting on the car
m = the mass of the car
a = the acceleration of the car


If you want to increase acceleration, you need to reduce weight, increase thrust, or decrease other forces that oppose thrust (namely drag).

Torque
Torque is a force applied about a moment arm (commonly described as a “twisting” force). Instead of a force that propels an object forward, think of it as a force that causes the object to rotate. If I apply a force about some point, the torque applied is the product of the distance to the point and the portion of the force which is perpendicular to that distance. The equation used to calculate torque is:

τ = Fd
Where, τ = torque
F = the portion of the total force acting perpendicularly to the distance vector
d = the distance between the fulcrum and the point at which the force F is applied


Note that torque is instantaneous. It doesn’t matter how long or the distance over which the force is applied.

Work
Work is the energy used to move an object. Forces and torques might always exist, but unless the object has moved, no work has been performed. If I try to move an object, but I can’t make it budge, then I’m performing no work. I’m using a lot of energy to develop force, but it’s being used to create heat, not to perform work. Once the object moves, then I’m using some of that energy to perform work. The equation used to determine the work performed on an object is:

W = Fx
Where, W = the work performed
F = the force applied to the object
x = the displacement of the object


Work can also be performed on a rotating object, such as a crankshaft. The equation in this case is slightly different:

W = τθ
Where, W = the work performed
τ = the torque applied
θ = the angular displacement (in radians)


Instead of force, torque is used, and the displacement is expressed in radians instead of feet or meters. If an object makes one revolution, its angular displacement is 2π radians (or 360°), if it makes ¼ revolution, its angular displacement is π/2 radians (or 90°), etc. Based on this equation, it should be apparent that torque is not a measure of work. If torque is applied to a shaft, but it does not rotate, no work is performed. It’s a little confusing since both have the same units, but they do not measure the same quantities.

Power
Power is a measure of how much work is performed (or energy used) per unit of time. It doesn’t measure how much torque is applied over time, just the work performed. For engines, the unit “horsepower” is used. If an engine produces 1 hp, then it can perform 550 ft-lb of work (not torque) every second. It doesn’t matter if the engine produces 1 ft-lb of torque or 100 ft-lb of torque. The equation for power is:

P = Fv
Where, P = power
F = the force applied
v = the velocity at which the force is applied


For a rotating object, it’s:

P = τω
Where, P = power
τ = the torque applied
ω = the rotational speed at which the torque is applied


For engines, we commonly see the equation: Horsepower = (Torque)*(RPM) / 5252

However, what does this really mean? Where does the constant 5252 come from?

Calculating Horsepower
As mentioned, power can be calculated by multiplying the torque applied to a rotating object by the speed at which it rotates. However, the base unit for rotational speed is radians per second, and we normally measure engine speed in revolutions per minute. Therefore, if we want to use rpm, we need to convert. Since there are 2π radians in one revolution, and 60 seconds in one minute, we have to divide by 60 and multiply by 2π:

P = (Torque) * (RPM) * (1 min / 60 s) * (2π rad / 1 rev) = 2π*(Torque)*(RPM) / 60

Next, we want power expressed in horsepower, not ft-lb/s (which is the base unit). Therefore, we need to divide by 550:

Horsepower = 2π*(Torque) *(RPM) / (60 * 550) = 2π*(Torque) *(RPM) / 33000

Since 33000/2π is about 5252, we just use: Horsepower = (Torque)*(RPM) / 5252

Now you can see that it’s not just a random equation with a meaningless result. It’s a measure of the rate at which energy is produced and the rate at which work is performed.

So What?
By now, you’re probably getting tired of my physics lesson, so I’ll move on. How do we use this information to design our engines? Do we want to produce tons of torque like a diesel engine or gobs of horsepower like a motorcycle engine? Let’s take a look at some examples.

I did a little searching and found a dyno plot from Race Prep Engineering which contains results from three different engines: a b16a, a d16a6, and a b18a. I digitized the curves and recreated them for you:





As seen in the graphs, the b16a produces 100.3 ft-lb of torque and 140.1 whp, the d16a6 produces, 104.5 ft-lb of torque and 110.9 whp, and the b18a produces 118.1 ft-lb of torque and 123.3 whp. As expected, the b16a makes the least torque but the most horsepower, the b18a makes the most torque, and the d16a6 makes the least horsepower. These curves are useful by themselves, but they don’t tell us anything about thrust, which is what we really care about. So, we need to do a little math. First, I needed the gear ratios for the transmissions used with each engine:



I’ll also assume that each engine/transmission is used with a 23.4” tire. Using this information, and assuming perfect launches and shift points, we can figure out how much thrust each setup will produce at any speed.

However, there’s a catch. At low speeds, there may be too much thrust for the tires to handle, and they’ll spin. If the tires are spinning, they cannot develop as much force as if they were rolling (or slightly slipping, which is most desirable). Therefore, the car cannot use more thrust than its traction limit. Without going into too much detail, the traction limit of a car (at steady-state) is dictated by its center of gravity, weight, wheelbase, and tires. When the car is launched, these factors affect how much weight is transferred from the front wheels to the rear wheels. However, FWD cars want as much load on the drive wheels as possible, and weight transfer works against them. This puts them at a major disadvantage for drag racing. I’ll spare everyone the math, but in this example, I’ve estimated that the traction limit is about 1344 lb for a typical Civic using street tires, which corresponds to a maximum acceleration of about 0.54 g.
 

· Registered
Joined
·
4,359 Posts
Discussion Starter · #2 ·
Maximum Tractive Effort
When all of the above information is combined, we get a plot like this:



In this chart, you can see how much thrust each engine can produce at any speed. Notice that below 30 mph, all three engines can only produce 1344 lb of thrust, since the car is traction-limited. You can also see how thrust drops five times in each curve – once at the end of each gear. Based on the results, the b16a produces an average of 858.3 lb of thrust between 0 and 120 mph, the d16a6 produces an average of 774.5 lb, and the b18a produces an average of 821.8 lb. Since this comparison uses the same drag and weight for each case, each car’s acceleration will follow these trends. That means the b16a can accelerate (on average) 4% faster than the b18a and 11% faster than the d16a6.

Why does the d16a6 make so much less thrust?
Let’s take a closer look at the curve for the d16a6. You’ll notice that the x-axis is vehicle speed (velocity), and the y-axis is thrust (force). If you remember from earlier in this article, the product of velocity and force is power. If the engine can only produce 110.9 whp (as is the case for the d16a6), then the product of vehicle speed and thrust can never be greater than 110.9 horsepower. Whereas engine torque is manipulated by gearing and tire size, horsepower remains constant (neglecting drivetrain friction). Now, let’s plot the curve for the d16a6 alongside its horsepower limit. I’ve included the b16a curve for comparison:



You can see that the curve for the d16a6 touches its horsepower limit five times – once in each gear. These points occur when the engine is producing peak power, and for any given speed, that’s when the car will accelerate fastest. Therefore, if we can change the torque curve and/or gearing so that the tractive effort curve rides closer to the horsepower limit, we can produce more average thrust, and car will go faster.

What if the d16a6 has better gearing?
Fair enough. Let’s give the b18a and the d16a6 the same gearing as the b16a and see what happens:



The better gearing helps – but only by so much. The b16a still makes 3% more average thrust than the b18a and 9% more average thrust than the d16a6. What if the b16a uses taller gearing and the d16a6 uses shorter gearing?



In this case, the b18a is close to the b16a, and all else being equal, it would only be 1% slower. However, the b16a still makes 7% more average thrust than the d16a6. It doesn’t matter how you gear the two cars; the d16a6 suffers by producing 30whp less than the b16a, even though it makes more torque.

The Bottom Line
If you want your car to be faster, you want more average horsepower. It doesn’t matter if you’re building a Formula 1 car, a Pro Stock car, or a weekend autocross car. In racing, horsepower is king. If you have a legitimate counterargument for this, feel free to share it. Otherwise, please stop saying “Horsepower sells cars, but torque wins races”.
 

· Registered
2001 Z3 Coupe 3.0i
Joined
·
1,106 Posts
great post, and its a good argument for maximizing your potential traction, too.
 
  • Like
Reactions: ddd4114

· Registered
same as above
Joined
·
1,447 Posts
rep for keeping at a highschool level so people will actually read it. id say that argument closed me.
 
  • Like
Reactions: ddd4114

· Registered
1988 Honda CRX
Joined
·
8,413 Posts
That was a very accurate and detailed set of posts and I agree with everything it says.

However it skims past or overlooks or implies, but does not spell out several practical points.

Torque and power in a running typical car engine are directly related by the constant 5252, so at any speed, a change in torque as measured by a dyno is the same in direct proportion to a change in power. Increase torque by say 10%, you get an increase in power by 10% at that speed. A dyno actually measures torque and speed, not power directly. The power is simply calculated by the torque, speed and a constant at a number of rpm points and the power graph is derived.

assuming a finite number of gears and that changing gears wastes some time, the real measure of engine performance is area under the power curve which is directly related to the area under the torque curve. There is a point where to sacrifice a little peak power to save a gear change pays off OR a higher peak power with an extra gear change pays off, but the maths are to complex and tedious for me. Some computer programs can do that for you.

By the methods you used, there will be optimum ratios and gear change points for any car, weight, tyre, power band.

A wider power band makes it easier to get close to optimum with less gear changes.

If we are chasing maximum torque, rpm does not matter, we need increased cylinder pressure or increased capacity while retaining cylinder pressure. There are several methods that can do this but they may reduce the rpm potential.

If we are chasing maximum power, we need to increase torque or increase rpm or both. Most real performance modifications aim to do this. Mant increase power by increasing engine speed but reduce width of powerband.

Engines with a broad power band with higher average power over a reasonable rpm range will generally beat an engine with a very narrow power peak but lower average power over a reasonable rpm range.

The engine with a wider powerband will be easier to drive and probably more durable and certainly more streetable.

When people generally talk of power vs torque, in their ignorance, they generally mean more mid range power and wider powerband vs more top end power and narrow power band.
 

· Registered
Joined
·
233 Posts
really interesting . I like this post.
So in simple stupid terms ...
- Power is directly proportional to the torque produced.
- more thrust , the faster the acceleration
- even though the D16a6 produced more torque then the B16 ,the B16 as seen in the graph's create more thrust which is proportional to the horsepower made.
D16 produced less horsepower then the B-series engiens, the B series which produced more horsepower is able to create more thrust and in the end accelerate more faster allowing them to be more dominant in this case.

is that it? just making sure , also when I say thrust is proportional to the amount of horsepower produced is that a correct statement? so the more horsepower produced the more thrust is being produced which allows you to accelerate?

thanks. just want to make sure I understand this because it is a good discussion!
 

· Registered
1988 Honda CRX
Joined
·
8,413 Posts
Another follow on point to consider is that torque, not power breaks drive train parts, well at least to the final drive unit.

This means a low torqur high rpm high power engine can run a lighter gearbox and clutch and input shaft in the gear box. This makes the car lighter, and may then allow another follow on that allows the chassis to be lighter and so on.

Once we reduce mass, particularly rotating mass, we don't need as much power to produce the same rate of acceleration.
 

· Registered
Joined
·
4,359 Posts
Discussion Starter · #16 ·
really interesting . I like this post.
So in simple stupid terms ...
- Power is directly proportional to the torque produced.
- more thrust , the faster the acceleration
- even though the D16a6 produced more torque then the B16 ,the B16 as seen in the graph's create more thrust which is proportional to the horsepower made.
D16 produced less horsepower then the B-series engiens, the B series which produced more horsepower is able to create more thrust and in the end accelerate more faster allowing them to be more dominant in this case.

is that it? just making sure , also when I say thrust is proportional to the amount of horsepower produced is that a correct statement? so the more horsepower produced the more thrust is being produced which allows you to accelerate?

thanks. just want to make sure I understand this because it is a good discussion!
Yes, that's all correct. At any instant in time, thrust is more or less proportional to the power developed by the engine. Frictional losses are not constant, and they decrease the amount of energy that makes it from the crankshaft to the wheels, but that's a separate discussion. Some of the energy is also used to increase the speed of rotating drivetrain components. The easy answer is that they're proportional.

Pat made a very good point that I didn't explicitly say in the article. As one quantity changes (either torque or horsepower), so does the other. When people say "I want more low-end torque", what they really want is more low-end horsepower, since they go hand in hand. Although horsepower is still what's needed to accelerate cars, pull tractors, and tow trailers, the engine speeds at which most power is developed might be different depending on the application. That's one of the reasons why diesel engines are great for large trucks. They make a lot of horsepower at low speeds, so the driver doesn't have to rev the piss out of the engine to go anywhere. It also keeps them (relatively) quiet and more reliable.

And yes, a wider, flatter torque curve makes the car easier to drive and more predictable. Bone's dyno plot is a good example of that.
 

· Registered
Joined
·
79 Posts
ddd4114 said:
The Bottom Line
If you want your car to be faster, you want more average horsepower. It doesn’t matter if you’re building a Formula 1 car, a Pro Stock car, or a weekend autocross car. In racing, horsepower is king. If you have a legitimate counterargument for this, feel free to share it. Otherwise, please stop saying “Horsepower sells cars, but torque wins races”.
First off good write up! I think your perspective may not be quite aligned with what the point of that quote is though.

Which is the dependent variable in the power formula? Is hp not a function of torque and rpm? In other words no torque = no hp. In another light, more average horsepower = more average torque. Your are essentially reinforcing the fact that torque is the key factor deciding "quickness" of a car.

The real lesson in all of this is: Peak numbers dont mean shit... Its all about the highest average numbers. My 64' 390ci mercury makes 400 ft/lbs.. Is it fast?no... 400 x 2000 rpm does not equal much work done. Same goes for the car that makes 300hp in a 200rpm window... Sure it can do alot of work, but how long is it actually doin this work? How often is the car inside this 200rpm window? Dont get me wrong a huge hp number in a short rpm window can work (pro stock anyone?) but your gonna have to make goldy amounts of power in your small window, to outwiegh the work potential of a less power larger window car. Its not an arguement of hp vs torque, its all about who can do more work in the quickest amount of time. Hp= work, torue and rpm dictate work..

Of course this is all theorhetical... When actually on the track maybe because of the speed of the corners, the torque we are making at that speed is past the threshold of traction and we would be better off maybe rockign the curve towards the top end... Racing involves soo much more then tryin to make the quickest car, where and how that work is delivered is crucial.

Regardless when you say more average hp, you are in effect saying more average torque over the same particular rpm. The quote goes back to the hp dynno guys building engines for some flashy number when in effect, their build wouldnt be so hot in application, compared to the guy with the stable torque line, whos torque, or resulting hp, has a higher average work done.
 

· Registered
Joined
·
4,359 Posts
Discussion Starter · #19 ·
Yes and no. I'm still saying that you want the most average horsepower, regardless of torque. You could make 300 horsepower from 3000 rpm to 5000 rpm, or you could make 300 horsepower from 7000 rpm to 9000 rpm, and (all else being equal) with optimal gearing, both would be just as fast. It's much easier to work with the torque curve, but the horsepower curve is the real indicator of performance. It's all about production and conservation of energy.

That being said, I am implying that to go faster, you do need to make more torque in the powerband, since they are directly proportional. However, it's the shape of the torque curve that's more important. Either make the same average torque at higher engine speeds, make more torque in your current powerband, or make a lot more torque lower in the powerband. However you get more average horsepower is up to you, but that should be the goal if you're looking for more performance. You just need to take driveability into account, as well as noise, reliability, etc. if it applies.

I do agree that peak numbers don't mean shit - especially if you can't use that much power. If your torque curve jumps from 100 ft-lb to 250 ft-lb in a 500 rpm window, and you're breaking the tires loose whenever you accelerate, you'd probably be faster with less power and a better overall curve. The tires are the four most important parts of the car, and if you're not using them to their full potential, there's room for improvement. Every performance aspect of the car should be designed around them.
 
1 - 20 of 142 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top