Why is 'driving with load' better than constant speed

[bikenfool]

Originally Posted by Old Mechanic

Going slower will increase mileage. That is a separate calculation.

Driving in hills is also a separate scenario.

On a dyno when you increase the load on an engine it will increase the power far in excess of the additional fuel required. 20 HP using 1 unit of fuel, while 50 HP uses 1.5 units of fuel. The additional 30 HP only cost half as much fuel as the first 20, so you are getting much more power for only half as much fuel. This overwhelms the aero drag calculation by a significant amount.This is due to the fact that the lower manifold vacuum under higher loads allows up to double the effective compression in the cylinder and much more power produced for the equivalent amount of fuel.

This is the essence of pulse and glide. My Fiesta will glide down to below 20 MPH and still get over 100 MPG.

Hills offer a chance to minimise the speed variations as long as the hill is not so steep that you can not maintain a safe, or legal speed on the downhill portion. The grade of the hill is crucial to the strategy of climbing and descending grades.

If your downhill coast can be maintained within a legal speed then you have the better choice. Lowest speed at the crest of the hill, highest speed at the bottom of the hill.
Climbing the hill allows you to store energy by increasing altitude. DWL alllows you to pick a load that minimises your mileage hit while climbing with maximum benefit in the downhill coast where you mileage can soar to hundreds of MPG. If you climb at 20 MPG and coast downhill at 200 MPG then it is easy to see that you can average excellent mileage overall.

If your peak speed downhill exceeds safety or legality then just use engine braking instead of brakes, since engine braking consumes no fuel, compared to coasting in neutral, but it is better to coast over engine braking in most cases, again depending on the grade of the hill.

While aero drag does increase as the square of speed, it does not increase to the point where the increased engine efficiency in the pulse overwhelms the gain in economy if you use the pulse and glide technique, as long as you do not exceed safe speeds and or legal speeds. It also does not mean you have to average a lower speed as long as you keep the variations in the range of your target speed.

Using this technique, refined over many miles of experience, I can average over 50 MPG in my Fiesta on the Interstate, while maintaining the average speed of traffic, in a car rated at 38 MPG highway, but the hills here are not steep enough to coast at 65 MPH. To achieve that mileage at that average speed requires focus, work, and concentration, and some drafting also helps.

regards
Mech

I understand what you're saying about increasing the load on the engine makes the engine more efficient. Perhaps you can clarify this: "DWL alllows you to pick a load that minimizes your mileage hit while climbing with maximum benefit in the downhill coast where you mileage can soar to hundreds of MPG." What is that load that you use? Efficient load is 1/2-full throttle. But most descriptions of DWL say to back off on the throttle. For most of the hills around here, to get up them in high gear I'm at a high load on the engine.
To be clear, I'm not questioning P&G. I still think DWL is a bit of a special case, or perhaps just misnamed.

[user removed]

Throttle position and load are two different things. Manifold vacuum determines load, higher vacuum=lower load. When you transition from downhill to uphill the load increases beyond the percentage of throttle position. Efficiency is measured by the actual compression in the cylinder when combustion begins is directly related to manifold vacuum and can vary significantly with very small changes in throttle position.

My Insight had an instant economy gauge which I used to refine my acceleration technique by accelerating with the lowest loss in instant MPG while providing an increase in speed which could then be used for a period of coasting. The Insight would coast in neutral at 150 (max measurement) in instant MPG down to 16 MPH which indicated and idle consumption of .11 GPH. When coasting at multiples of 16 MPH the instant MPG would be 150Xspeed divided by 16, so at 64 MPH the instant MPG would be 600 MPG. Using this on slight grades I could average over 90 MPG in the Insight and it was with a CVT transmission.

If you go by throttle position then at a point in load the AT will downshift and you will suffer a dramatid loss in instant MPG. Without precision instrumentation the best technique is to avoid any downshifting in a auto transmission to the point where you are loosing to much speed (when climbing uphill). This is where you would wantt to consider allowing your speed to slowly drop in order to maintain top gear. The same also works with manuals. Avoid downshifting to minimize the loss in MPG climbing the hill, but not to the point where your speed drops below about 75-80% of your desired average or you take so much longer to get to the top of the hill at lower instant MPG.

I will even accelerate before reaching the bottom of the hill, if I risk to long on the climb.

As I said before it really depends on the hill and you strategy changes as the grades increase, but if you do it correctly you should do fine until the downhill portion does not allow you to coast without engine braking or any use of friction brakes. Only when the grades get beyond that percentage will you suffer loss of average MPG.

I call it the roller coaster scenario. The coaster climbs the grade initially and from that point onward it is all inertia. The distance the coaster travels is many times greater coasting than the initial climb. I remember one member here telling about coasting 35 miles coming out of the Rocky mountains. My longest coast has been about 3.5 miles.
That was also enhanced by drafting on I85 coming east from Blacksburg Va. I had to used engine braking, in my Insight and the battery charge dropped dramatically on the lcimb but recovered completely on the downhills. I average 70 MPG with an average speed of 55 MPH in the Insight on that trip. I think it was 635 miles on 9.65 gallons of fuel.

regards
Mech

[user removed]

Best load is 75-80% with out full throttle enrichment. I will drive many tanks of fuel in my Fiesta and never see over 3500 RPM, becasue that is where it shifts out of first gear even when accelerating gradually. I find that when I am in heavy traffic it is actually best to accelerate at the average speed of the traffic and it also creates less issues with other drivers wanting to tailgate you if you accelerate at a slower pace.

regards
Mech

[bikenfool]

Originally Posted by Old Mechanic

If you go by throttle position then at a point in load the AT will downshift and you will suffer a dramatid loss in instant MPG. Without precision instrumentation the best technique is to avoid any downshifting in a auto transmission to the point where you are loosing to much speed (when climbing uphill). This is where you would wantt to consider allowing your speed to slowly drop in order to maintain top gear. The same also works with manuals. Avoid downshifting to minimize the loss in MPG climbing the hill, but not to the point where your speed drops below about 75-80% of your desired average or you take so much longer to get to the top of the hill at lower instant MPG.
regards
Mech

Originally Posted by Old Mechanic

Best load is 75-80% with out full throttle enrichment. I will drive many tanks of fuel in my Fiesta and never see over 3500 RPM, becasue that is where it shifts out of first gear even when accelerating gradually. I find that when I am in heavy traffic it is actually best to accelerate at the average speed of the traffic and it also creates less issues with other drivers wanting to tailgate you if you accelerate at a slower pace.

regards
Mech

If your RPM's and load are in an efficient zone, why slow down at all on a hill with a manual? If you slow down you risk having to shift down.

[user removed]

If you can not maintain speed in top gear then you are better off to downshift than go into full load enrichment with throttle at 100%, but it is close either way.So close that it would depend on your priority, higher average speed or higher mileage. A good rule is the opportunity for vastly better mileage is what you do on the downhill portion, not the uphill climb.

If you can coast downhill and your terminal speed will not exceed your safety or legality threshold, then you objective should be principally based on your speed at the BOTTOM of the hill. It's like thinking ahead in a Chess game. You want you speed at the top to be at the exact point where your speed at the bottom is the highest practical speed (legal and safe) and you have coasted dowhilll to reach that speed, while your mileage jumps to way over 100 MPG, in some cases over 300 MPG, or infinite MPG if you use engine off coasting, which I do not do in my driving.

Then you have more speed to help with climbing the next hill (assuming there is one). Maintaining the best top gear load will cause your speed to gradually drop as you climb hill #2 in order to again reach your terminal velocity at the bottom of hill 2. Rinse and repeat for additional hills.

regards
Mech

[euromodder]

Originally Posted by bikenfool

the optimum solution is constant speed.

That's the general thinking.
But it's wrong.
The most efficient is pulse (accelerate @ BSFC for your engine) and glide (engine off). Repeat and repeat ...
Even though your speed varies significantly, it's more efficient..
Constant speed means the engine is running inefficiently - even if its fuel consumption is relatively low (compared to what you're used to).

With Pulse & Glide, you use the engine at its peak efficiency whenever it's on.
That's what makes it so efficient.


Even coasting with the engine on (and driveline disconnected) will improve your efficiency.

I don't see how that comes into play for DWL or CT.

Constant speed is fine - compared to "regular" driving - as long as the road is level. Actually, you'd be DWL at that point.

But you lose efficiency whenever you go uphill or downhill.
Uphill, a CC or the driver simply pours on fuel to keep the speed up.
Downhill, it's actually using engine braking to keep the speed under control - and braking is not good. The car wants to go faster due to the grade, so let it go faster, if it can be done safely.

Slowing down on the hills saves fuel on the hill, but also takes longer. Wouldn't it be better to just slow down the average constant speed?

The slower you go, the better your average - usually.

But there's a point where the engine becomes more inefficient as you slow down, and you may get the same MPG as you would at a higher speed.
The benefit then is in not accelerating to the higher speed - which will be minimal on longer drives.

But even if the slower average is improving your MPG, it'd still be more efficient to slow down uphill and catch up downhill.

Perhaps its just a conservation of energy thing?

That's part of the point.

You trade fuel for potential energy : your car's weight @ height.
The engine load increases, and so does your engine's efficiency.
High load and low rpm is where it gets its BSFC after all ...
Even if it uses more fuel/mile, it's producing the horses required to get up there more efficiently.

Going downhill, you can trade height for speed (and distance) again.
But if you hold it back, you'd be braking and you won't get the return on your invested fuel.

you're gaining economy but losing time.

Despite slowing down considerably, I'm not losing that much time on my 30 mile commute.
On longer trips, the loss can be considerable, so there's usually always a trade-off between time and MPG.

If you accelerate with throttle on the descent I don't think you'll gain much, if any, in fuel economy.

You don't really accelerate with throttle when going downhill while using DWL.
You just get the car back to the speed it wants to make on as much fuel as you're giving it + the downhill grade

Some hills you can even glide down with no throttle at all - and driveline disconnected !

If you brake, you won't be getting as much of the potential energy (i.e. the fuel invested) back though.

Then when the road levels out, you'll be back to your DWL fuel target, the sort of fuel consumption you want to achieve .


Try DWL instead of CC for a few tanks.
It will give you improved MPG

[bikenfool]

Euromodder, I agree with most of what you're saying, but most of what you're saying is about P&G, not DWL which is what I question. Hills seem like P&G enhancement. Pulse on the climb, glide on the descent.
DWL implies constant load on the descent, in other words using the engine to increase speed. It's obvious gliding is best.

Originally Posted by euromodder

But you lose efficiency whenever you go uphill or downhill.

Uphill, a CC or the driver simply pours on fuel to keep the speed up.

I know you'd be using more fuel, but going uphill & keeping the speed up would usually be more efficient. When you're loafing along on flat ground your below the optimum region on a BSFC map. Higher load would bring you up and be more efficient.

Originally Posted by euromodder

You trade fuel for potential energy : your car's weight @ height.
The engine load increases, and so does your engine's efficiency.
High load and low rpm is where it gets its BSFC after all ...
Even if it uses more fuel/mile, it's producing the horses required to get up there more efficiently.

That sounds like a pulse in the climb, not DWL, they are opposing techniques.

Originally Posted by euromodder

Try DWL instead of CC for a few tanks.
It will give you improved MPG

I've been playing with it, but I don't typically use CC locally anyway.

[NachtRitter]

Originally Posted by bikenfool

DWL implies constant load on the descent, in other words using the engine to increase speed. It's obvious gliding is best.

When I read what you wrote, it seems to me that you are under the impression that a hypermiler sticks to exactly one technique no matter what the condition. What you seem to be trying to say is that when the hypermiler crests the hill, he (or she) will press on the accelerator to maintain a specific load ... when legal speed is reached, the hypermiler will continue to press on the accelerator to maintain that specific load ... when a sharp curve is approaching, the hypermiler will continue to press on the accelerator to maintain that specific load ... no, that's just silly. The hypermiler will do what he or she knows how to do or is able to do which fits the current conditions. That means DWL up the hill if that's appropriate for the conditions and then switching to a different (appropriate) technique (e.g. coasting, engine on or off, in gear or out of gear) when going down the hill. This "switching of tactics" does not violate the description of DWL given on this site. You can read whatever you want into it, but apply some logic to your assumptions.

Originally Posted by bikenfool

I know you'd be using more fuel, but going uphill & keeping the speed up would usually be more efficient.

Please explain your reasoning here. How will using more fuel per mile on the uphill get you better fuel economy, keeping in mind that you know you have a downhill once you've climbed the uphill? In your scenario, how would you maintain your speed within legal limits on the downhill? Engine braking? Brakes? Somehow, the net amount of fuel consumed on the uphill, downhill combination needs to be lower than technique of using DWL on the uphill (possibly losing speed) and then coasting on the downhill (gaining the speed back "for free"). Consuming more fuel on the uphill than one would consume using DWL won't net you better FE overall.

[euromodder]

Originally Posted by bikenfool

Hills seem like P&G enhancement. Pulse on the climb, glide on the descent.

The climb can be seen as a pulse (but in potential energy, not kinetic / speed), though you'd still be wanting to slow down uphill.

You usually need to carry some speed though
There are situations where you can crest a hill or bridge with very little speed, and catch back up downhill. Traffic allowing.


Sticking to the same speed while going uphill is already a fuel drain.
You really don't want to have to accelerate uphill, it's a double lose-lose situation.

DWL implies constant load on the descent, in other words using the engine to increase speed.

Regain the speed you had, and let gravity add a bit on top of that, recovering some of the energy wasted going uphill.
The extra speed will then slowly wear off once you're on level ground.

I've found this to be a condition that engines respond quite well to: slowly letting speed decay a wee bit, and fuel consumption drops, then stays low at the slightly slower constant speed.


As you DWL uphill, engine load increases while rpm are reduced, this both boosts your engine's efficiency
(as long as it can make enough power at those rpm - if it can't, you have to shift down)

I know you'd be using more fuel, but going uphill & keeping the speed up would usually be more efficient.

It's not, and it's definitely not the case when it means the higher cresting speed means overspeeding (going beyond the posted or your own limits) on the downhill stretch.

While you're making the HP efficiently at the high load required to go up, the HP required means you also burn more fuel - even if it's done efficiently.

Requiring less HP to start with, but still in an ara of high engine-efficiency, translates in requiring less fuel.

When you're loafing along on flat ground your below the optimum region on a BSFC map. Higher load would bring you up and be more efficient.

When loafing along on flat ground, you're certainly well below the BSFC ...
Higher load would make the engine more efficient, but it'd also use more gas.

You got to disconnect the engine's best fuel-to-HP efficiency from the overal fuel efficiency.

At best BSFC, you're typically using a lot of gas - yet doing it so efficiently to make a lot of HP, that you can use it to save gas overall (i.e. using Pulse & Glide)

At your best fuel consumption speed, you'd be well below the best BSFC of your engine, so your engine isn't using the fuel efficiently, but it needs so little HP to sustain level constant speed that you're still using only a little bit of fuel.


A vehicle's engine is sized for peak performance, but using DWL or CC, it typically runs quite inefficiently (@ converting fuel to HP) at low average power output to sustain a constant, level ground, speed.
The low HP requirement is really the only saving grace in this situation.


That's what makes P&G so efficient :
When it needs to make power, it's done at the best fuel-to-HP conversion rate.
When it doesn't need power, the engine is off , not using any fuel.
The average power requirement is the same for the same average speed.
But with P&G, this power is made far more efficiently when needed.


It also means DWL will be less fuel efficient overall than P&G.
With DWL you mostly run the engine at low power requirement, low load, but that's where it's not running efficiently @ making HP.
It's only needing so little HP it uses little gas.


The really astonishingly high MPG results, are most often achieved using P&G.
It's a shame really that P&G is a high workload driving technique, and not always compatible with other traffic.
It'd drive other ppl mad around here

[bikenfool]

Enough talk, lets see some analysis.
Lets look at 2 cases. Perhaps someone can compute the fuel consumption for each case with a typical car. Manual tranny, non hybrid.

The road is a long flat section, followed by a climb. The time to cover the course has to be the same for each method. You can pick the parameters for the length and grade. Try a couple of different courses. The intial and final conditions are the same for each method.
Method 1. DWL. It would be way simpler to do this in top gear, no down shift.
Method 2. Constant speed, begin a coast before the top of the hill so the speed at the top is the same as for the DWL case. This takes the descent out of the equation since you could use the same method (glide) for either case on the down side. The speed on the flat section will be slower for this case than for the DWL case since you don't lose time on the climb.

I know that you can save fuel by arriving at the top of the hill with a slower speed, but there are other ways to accomplish that such as the glide I just proposed in 2)
I know that you can save fuel on the climb by going slower, but you can also save fuel by going slower on the flats, so there is a trade off.

Maybe I'm overthinking this, but I don't think the solution is as obvous as you guys think it is.

My brief search didn't show up any open source simulations, but it seems like there's probably something out there somewhere. Perhaps some government project. It wouldn't be trivial, but it wouldn't be a huge task to code something up in matlab or octave. If someone knows of a sim out there point me to it. It doesn't have to be very complicated. This simple case could probably be done with some fairly simple script, but I won't be able to try it for a while. I'm sure one of you guys already have the tools ready to go