New here, ... the physics of hypermiling?
 

(reposted from CleanMPG...)

I've started hypermiling in my '93 Mitsubishi Expo LRV, 1.8 L 4G93 SOHC 8-valve engine. Multiport Fuel Injection. No instrumentation (yet, can't decide). Just fixed the speedometer cable, broken for years. Hey, it runs great without it! (Money is tight)

I'm new to all this, looking at a lot of information about FE , etc. but I'm coming to a few simplifications to understand hypermiling. Tell me if this is right. (I was a physic major so... it guides my thinking).

Basically, I'll get the best mileage if I either have the ICE off, in a coast (or because I took my bike!). Or, I'm operating at minimum BSFC, probably rapidly converting gas energy to kinetic energy (accelerating for later coast) since all cars are in a sense highly overpowered. Guessing thats somewhere at 75% throttle, 2000 rpm. We have a sledge hammer to drive a tack. The big trick is then not getting arrested, killed by road rage, crashing, or having to hit the brakes wasting that kinetic energy. But last night I came all the way home without touching the brakes. I mostly bump start, brief pulse at high throttle about 3-5 sec, ICE off, long coast. No idea of my mileage.

Other FE effects: weather, choice of route, managing engine warm-up costs, good maintenance, having "efficient hardware" to start with (your choice of car: engine, aerodynamics, rolling resistance).

There's a question of choice of gear and average speed (in theory always running at minimum BSFC), but that's kind of a trade-off of time to destination .vs. fuel used (i.e. time is money) because wind drag increases with (the square) of speed. We decide how much our time is worth .vs. fuel. Also, there's limited gear choices and manual or automatic control.

Where is the minimum BSFC operating point (s)? It seems it would be helpful to have a procedure to map this out practically, then instruments to tell me where I'm at, say with a bar for rpm and throttle position, above or below exact minimum. Rather, it seems most people "hunt" for it with data clouded with all other effects, tracking instantaneous and averages of FE. I'm confused what instrumentation to get. I'll post separately all that.

In general, what features, describe an (market available or not) car best for Pulse and Glide as I describe it? (apart from "efficient hardware") A manual transmission with kill switch? Or turning key-in-ignition Fas ? Some make/model hybrid? What size engine would it have? Once you completely put your driving behavior on the table to the point of advanced Hypermiling techniques, it seems current hybrids are not quite right. They're better for people who don't want to change driving behavior (so much). Or think about it. Of course, the "efficient hardware" tends to be in hybrids though.


Allan

Sounds like you have a good grip on all the stuff that is the foundation of hypermiling.

As far as gear choice and speed, I think you want to be in top gear as often/as long as possible. For me, if that means a lil bit of speeding in town, so be it (ex: In an a/t car going fast enough to get converter lock-up if I think it won't result in a ticket. BTW, I've found you can back off up to 5 mph from lock-up before it unlocks again) and conversely out on the highway if it means being a lil under the limit, OK. In m/t cars I'll often skip-shift. I tend to ignore my time or the value of it as far as speed selection goes since A) I don't monetize my time and B) In all but the longest trips, the time effects of speed are quite insignificant anyway.

As far as "best car" choice, you can look at what the top performers are driving- Civics, Metros, Priuii, and the like. IMHO people who do more metro-area driving can wring more out of hybrids while people who do a lil small town driving/ more open highway cruising are better off without hybrid expense and complexity.

Not really, unless I have this wrong. Rather, you should be as close to the minimum BSFC as possible. That's the best conversion of gas to kinetic energy.

I looked up my gearing and calculated RPMs:
Final diff ratio 4.021
Tire dia 25.3"
Mitsu has 4G93 engine and mitsubishi F5M22 transmission

If the min FC in the BSFC map is at 2000 rpm at 75% load, like I'm lead to believe, then I should be in 3rd at 30 mph, 4th at 40 mph and 5th at 50 mph, all at 75% load. Aerodynamic drag should be higher at 50 than 30, so mileage should be best at 30, not 50. But maybe I'm "in a hurry".

Unless you mean cruising at a steady throttle and 30 mph. COmpletely different. Then, yes, 5th at 1212 rpm would probably be better FE than 3rd at 2059 rpm. But it would be much worse than P&G above.

Do I have this right? Practically, though, can you really P&G for 12 straight hours, ICE off, on a trip of 600 miles? It begs for a better way.

It's all dangerous and illegal and stuff, but conceptually, I'm thinking of a mod than automates P&G via a on-off button on the steering wheel. Might be ok for lonely country roads.

Interesting, true about hybrids in stop and go.

It's hard to interpret results because I have no idea exactly how and where they are driving. Mostly how.

For BSFC, you need a B(rake). Anybody rigged an onboard torque monitor?

If P&G were at a high enough frequency, it would be unnoticeable, nondisruptive and safe.

And one more variable for the calculations: Everytime you kill the engine, you throw away the rotating kinetic energy in the engine. You could be better off getting real proficient at DWL and DWB.

Good data beats 1000 hypotheses.

There's probably some value to the kinetic energy of the rotating engine. HOWEVER remember you need to provide fuel to keep that thing turning. That's where you get a big benefit from moving with the engine off, as long as it's safe to do so.

Reciprocating pistons and loading/unloading the valve springs etc. all require some energy, not to mention the alternator, water pump, power steering, and the fuel pump that all will likely keep running unless you've put them on vacation.

My Civic, before doing the EPES (aka alternator-delete), needed right around .18 gallons per hour to idle at 670 rpm, fully warmed up. x 3 = .54 gph for 2010 rpm which would be maybe 55 mph in 5th gear?? If you're getting 55 mpg at 55 mph, that's 1 gallon per hour. And .54 gallon of it is to keep the engine turning!

My point is that you are using significant quantities of fuel just to turn the engine. That's why some of us are very keen on killing the engine for coasting.

My MPGuino reported that at last weekend's Green Grand Prix, I drove just about HALF the distance with the engine off. Usually these days I get low 50's mpg with the engine off may 10-15% of the distance. But at the GGP I scored 62 mpg.

How interesting.

I just started this, and haven't got any read on mileage, but I'm guessing I'm engine off about 80% of the time. If I'm not pulsing, it's off for the most part. Waiting to turn in the intersection, dodging traffic, no.

Is it just ... easier driving that determines 10-15% off? And half if ... someones looking? :-)

Allan

Thanks for the chuckle - easier driving and less EOC if someone's looking!

I think it has a lot to do with route. Or maybe I'm just not EOC'ing enough???

My commute is mostly highway, about 55 miles each way, with generally light traffic. EOC going down say 45 or 50 mph just isn't feasible if I want a reasonable commute time. My car coasts pretty well but even so, I need to run the engine most of the time to keep the car going at highway speeds.

Driving my HX I have one consideration most drivers don't have. The HX will run in lean burn mode, with the air:fuel ratio at something like 20:1 or maybe even more. I figure I net an extra 10-15% mpg when it's in lean burn. BUT if I kill the engine it won't do lean burn until at least 1.5 miles after restarting, and often it requires 2 miles to go into lean burn again. So usually I prefer to keep it in lean burn as much as possible, and let it idle in neutral if I can coast down a hill. I wouldn't be doing that idle-coast if it didn't have lean burn capability.

So at the GGP, which has a 2.45 mile loop of track, I did tons of EOC. That track has many hills as well as banked curves, is perfect for EOC. Not good for lean burn in the HX because you're constantly on and off the gas pedal; it would hardly ever go into lean burn under those conditions and would go right out of lean burn at the next upgrade or coast anyway.

The rule of thumb that I've heard is if the engine can be off for 15 seconds or more then it is likely better to turn it off. The one time that I coast a long time but leave the engine on is if the engine is still cold -- you need to bring it up to about 165F in my car to get it off of the richer mixture, so idling while coasting to warm it up seems worth it.

And bump starting is probably better than using the starter, so if you have a standard, you can turn it off, wait a second and then turn it on again so that you can bump start it as quickly as possible, as needed.

As you probably know, warmed up, newer cars with oxygen sensors go "closed loop" where fuel injection is adjusted based on the oxygen in the exhaust. When cold, they run "open loop" and inject an amount from the table... the "fuel map" (with actually many other adjustments). Lean burn is yet another mode but that's newer technology.

I'm under the impression somehow that wear and tear is high on a cold engine, and in Minnesota here,... sometimes that's 40F and sometimes that's minus 40F. I wonder if hard pulses on an engine not at operating temp is wasteful in terms of excessive wear.

Agreed. It seems safer too. Always do this, unless I get stopped at a light or something.

I take a slightly different view on warmup in very cold weather. My preferred would be to have a good electric preheater that warms the block via the coolant. I don't have one so I have to deal with warmup in New England winter, which I admit isn't as frigid as the Minnesota winters.

Anyway, the engine runs rich when cold, we've already discussed that. So if it HAS to run rich, I at least want to get the most possible distance out of the fuel it burns while warming up. No warmup idling for me, except maybe if it's 10 degrees or less out there and I don't feel safe "using" the engine at those low temperatures.

So while warming up, especially in winter weather, I pulse and glide as aggressively as I dare. I really avoid idling whether rolling or not. I also avoid coasting in gear; that mostly slows the car down and doesn't do anything much to warm it up. I'm not a big fan of wide open throttle, or even of 60-80% open throttle. So I'm using moderate throttle when engine is running.

I use a grill block. That really helps speed the warmup process, and also enables the engine to reach and maintain full warmup temperature even in frigid weather. In winter I'd say about 90% of my grill is blocked off. Mine is removable chunks of gray foam pipe insulation stuffed into the openings that are about 2" x 5". The foam is layered up and bound together with duct tape. I'm working on something prettier but the nice thing about this setup is that I can add and remove chunks as needed.

I pay attention to the engine temp display on my ScanGauge. When the outside temperature changes significantly I need to add or remove grill block chunks.

Thanks for commenting on this.

Yes, I'm thinking that would have a lot of benefits and self-installed, they are only about $50. You just wouldn't want to run them all night like we used to when I was a kid.

The grill block is very effective I think.

So you start EOC right away? I'd think full temp would take a good while at say 35 on city streets.

It would be as light as possible, very aerodynamic (both low frontal area & Cd), and have very low rolling resistance. All of which translates to less fuel required to accelerate (or climb), and less aero & rolling drag in order to permit the longest possible coasting distances. The engine power would be such that you just have "acceptable" acceleration rates (open to different interpretations) at high load/low RPM (minimum BSFC).

The mileage competition vehicles embody these goals to the extreme, eg:

2,757.1 MPG at Shell Eco-Marathon Americas
http://ecomodder.com/blog/wp-content...al-univers.jpg

As for a road-going passenger car, for minimum fuel consumption in a pulse & glide scenario, find (or make/modify) one with characteristics as close as possible to these types of competition vehicles.

Some hybrids are great for hardcore hypermiling because they are lighter, more aerodynamic, and have lower rolling resistance than comparable non-hybrids. Pulse & glide is a pleasure to do in a hybrid vehicle which is designed so you can control whether the gas engine is on or off and whether you are effectively coasting in "neutral" all solely through accelerator pedal position.

A small, light, aerodynamic vehicle with a manual transmission and a kill switch is the best non-hybrid alternative.

Didn't #3 get 12,xxx mpg "equivalent"?

Frank - could be. That one's from 2009 - I just grabbed it from the blog as an example.

Has anyone ever discussed a microcontroller that controlled the clutch, injection and throttle allowing for say, a "pulse/glide" button (unkill, engage clutch- throttle up to 75% until speed is +3 mph and then "glide" (throttle down to idle, disengage clutch, kill until speed is -3 mph, repeat)? Like someone just said, if it were cycled quick enough, to perhaps maintain speed +- 3-5 mph it would "pass" as "normal".

I'm actually kind of shopping for a replacement car. An engine swap is rather involved (but possible for me).

Any suggestions for a small engine and car? Loooowwww accelleration. I'm wondering what possibilities this has. No idea but guessing easy (bolt compatible) options are limited, and unplanned combinations involve custom machining and complexity.

Yes, I start EOC right away. And true, reaching full warmup would take a while on city streets, especially in a Minnesota winter. My house is only .72 miles from the interstate on-ramp so I'm in a different situation from what you're suggesting.

BUT you will get it warmed up a lot sooner if you set up an effective grill block.

Regardless, I'd start EOC as long as it's not so d**n frigid out that you're afraid to do anything but let it idle. We get plenty mornings in the 20's ºF here and I get rolling within seconds of startup, and EOC as soon as I'm not going uphill.

As I wrote a few posts back, the car is going to run rich till it's warmed up. No way around that. If it IS going to run rich, I at least want to get the max possible miles out of that quantity of fuel. So I'm going to EOC and whatever else I can think of to maximize it. Letting it burn more fuel than needed does make it run better once it's warmed up, but you're not getting any distance for that fuel you've burned.