"Square" Engine Theory

[spectre6000]

I've been trying to design an engine around fuel economy, and I ran into a few articles about Jon Karcey's Ghia. He used 77mm pistons and a 76mm stroke. HVW talked about building the engine "square" for the Mileage Motor articles (probably because Karcey did it). What's the logic there? Why is "square" a good thing? What does it do that a combination with a shorter stroke relative to the bore doesn't? What would happen if this relationship were turned on its head? I'm looking for some theory, not "my car has this and does that" anecdotes. Thanks.

[hotrodsurplus]

Sorry for the lack of formal intro. I post every now and then to the off-road forum but a friend forwarded this to me.

As far as 'square' goes, there's no real magic or merit to it. The person who's best explained bore/stroke/rod ratio relationship in the recent past was David Vizard. I can post some of his theory if you're interested. It pretty much explains the mechanics behind these often misunderstood specifications. It also debunks a lot of urban legend, specifically how a short-stroke engine is better at making high-speed power and a long-stroke is better at low-speed torque. Well, there IS some merit to it; a shorter-stroke engine of comparative size will have a larger bore and a larger bore will accommodate larger valves which will potentially flow greater volume which will promote operation at higher engine speeds. Make sense?

For the most part Karcey (and most other manufacturers) do small-bore engines for economy. The reason is that a smaller bore has a smaller surface-to-volume ratio. A specific cylinder volume with a smaller surface area (smaller piston face, reduced cylinder-wall area, and so on) theoretically requires less fuel to operate. In an ideal world the quench area forces all of the fuel mixture into the chamber, but in application that's not really the case. There's always a bit of unburned fuel in the quench area, so to give the cylinder the fuel volume it needs a larger cylinder will require more than a smaller cylinder will, even if the engine displacements are the same.

Naturally small-bore engines are limited to smaller valve sizes that limit top-end breathing, but engines built for high efficiency at lower speeds aren't concerned with high engine speeds. That's not to say a small-valve engine can't breathe; a relatively high rod ratio will make the most of an available small port and valve. A high rod ratio also increases dwell time at TDC, which is also a good way to increase efficiency and therefore potentially fuel economy.

But now I'm reading too much into it. As i said, I can post some of Vizard's theory on it. It makes sense of a lot of these things. I'm probably not going to make a whole lot of sense because I have about four slugs of rum in me right now and I'm pretty much three sheets to the wind. And I'm also not an 'expert.' I'm just a half-drunk guy who knows more than he can explain.

[Air_Cooled_Nut]

Just to add to the mix, here's a side-note I kept from a car magazine:


http://www.aircoolednut.com/erkson/temp ... square.jpg

[hotrodsurplus]

Here's a link to the instructions/foreword that David Vizard wrote for Desktop Dyno software. Yeah it's old(er) but the information is valid. Search (Cntrl F) for the subhead Fallacy Two: Long-Stroke Torque Vs. Short-Stroke Horsepower

http://www.nightrider.com/biketech/down ... _guide.pdf

It dispenses with the myth that long-stroke engines are better at torque building than short-stroke engines. Yes it goes against everything that 'everybody knows' about engines, but there are LOTS of short-stroke Torque monsters, specifically for us the Type IV engine (it has a shorter stroke than a 1600). There are other examples, too, like GMC's 60-degree v-6 engines of the '60s and early '70s. They have GIANT bores and teensy-tiny strokes and as a result suck fuel with impunity--and I should know because I own one and it's miserable.

This piece explains things in pretty great scope. Even though it's geared for people who want great HP numbers, it has a LOT of information that you can apply in building a mileage engine.

[spectre6000]

I would love to see the theory. That's what I'm after. I'm in the process of designing an engine, and this is one of the few things I don't yet feel that I have a complete grasp on. I get that a higher surface area gives the expanding fuel more to push on, and it seems that this would result in a more efficient operation. I don't necessarily get why though. I'm sure it's fairly simple, just not readily apparent for someone without a mechanical engineering degree.

Also, I know the theory behind rod ratios, and this is a slight change of subject, but short of having a number of different sets of rods available for mocking up and taking measurements, how would I go about determining the longest rods that will fit in a given setup? Specifically a late 1300 (1600 case w/ 77mm cylinders) w/ a 76mm crank.

[hotrodsurplus]

I get that a higher surface area gives the expanding fuel more to push on, and it seems that this would result in a more efficient operation.

Actually not necessarily so. The increased surface area will actually hurt fuel economy because it requires more fuel to saturate the surface area properly compared to a smaller bore. Yes, the larger surface area offers more area on which to 'push'; however, remember that the increased stroke in a smaller-bore engine of the same size increases the leverage on the reciprocating mass. So in effect it's a wash.

I don't necessarily get why though. I'm sure it's fairly simple, just not readily apparent for someone without a mechanical engineering degree.

Yeah, it's a LOT of information to learn. I commend you for doing your homework, though. Not a lot of people do; they just copy what they've seen in magazines. That Vizard article has some really good info in relatively plain-spoken english. Then again, you'll have to read it several times before it sinks in. I knew a lot of the information beforehand and I STILL had to go through it several times.

short of having a number of different sets of rods available for mocking up and taking measurements, how would I go about determining the longest rods that will fit in a given setup? Specifically a late 1300 (1600 case w/ 77mm cylinders) w/ a 76mm crank.

Here's a very fortunate combination. A Chevrolet 5.7 con rod on a 74mm crank gives you a 1.95:1 RR, which is just about dead-nuts with the stock Type I rod ratio. The additional 2mm stroke for a 76 crank would reduce RR to 1.9:1, which is a negligible drop. You'd never ever miss the .5 spread.

People will say that the reduced rod journal size will make the crank weaker (which is theoretically true); however, you won't be spinning this thing fast enough to reveal any problems (I've had longer-stroke engines with Chevy rods that lived good and long even at high engine speeds).

That 5.7" Chevy rod is the economical alternative. Any longer a rod requires custom work and that's big bux. The 5.7 rod with that 76 stroke will fit easily, even if you're forced to use a piston with a stock pin height.

[spectre6000]

I get that a higher surface area gives the expanding fuel more to push on, and it seems that this would result in a more efficient operation.

Actually not necessarily so. The increased surface area will actually hurt fuel economy because it requires more fuel to saturate the surface area properly compared to a smaller bore. Yes, the larger surface area offers more area on which to 'push'; however, remember that the increased stroke in a smaller-bore engine of the same size increases the leverage on the reciprocating mass. So in effect it's a wash.

Right. I Misspoke. The increased stroke is in relation to to the bore. How is it a wash though? Certainly a reduction in bore with all else staying the same will not result in a complete wash, you must mean in relation to some specific variable.

[spectre6000]

Here's a very fortunate combination. A Chevrolet 5.7 con rod on a 74mm crank gives you a 1.95:1 RR, which is just about dead-nuts with the stock Type I rod ratio. The additional 2mm stroke for a 76 crank would reduce RR to 1.9:1, which is a negligible drop. You'd never ever miss the .5 spread.

People will say that the reduced rod journal size will make the crank weaker (which is theoretically true); however, you won't be spinning this thing fast enough to reveal any problems (I've had longer-stroke engines with Chevy rods that lived good and long even at high engine speeds).

That 5.7" Chevy rod is the economical alternative. Any longer a rod requires custom work and that's big bux. The 5.7 rod with that 76 stroke will fit easily, even if you're forced to use a piston with a stock pin height.

Is there any reason to prefer a Chevy journal over a VW journal if both are available? I'm led to believe a 76mm crank is the longest stroke that will fit a standard case without any machining and is available in both Chevy and VW journals sizes. I've found some 5.7" rods with Chevy journals with a description suggesting they'll fit a standard case without any machining up to a 76mm stroke. I'm sure I can find the same in VW journals with some searching.

[sideshow]

Is there any reason to prefer a Chevy journal over a VW journal if both are available?

Sometimes a "big cam" will only clear if you limit the size of the crank. So rod to cam clearance, and to a lesser degree rod to case clearance,

[Ol'fogasaurus]

One additional thing to think about is a long toss (stroker) also has the increased piston speed you have to deak with, This is because of the longer distance the piston and rings have to travel over a standard short stroke engine. Just throwing this in the muddy the water.

Also the design of the cam shaft is so important in making a grunt/torkey vs. a hp engine. More lift with a short duration can produce power. Add to that properly sized valves, combustion chamber plus intake and exhaust runners design both in the head and manifold can make a big difference in the hp vs. torque discussion. Big isn't always beautiful.

[spectre6000]

The plan, assuming the square engine logic has merit, is to get a set of 1300dp heads (77mm bore) and run them with a 76mm crank. Additional heat from the extra friction generated by the faster/longer piston travel will be offset by the use of a set of Nickies. I want to make sure the "square" thing isn't just marketing hype. Karcey got some impressive results and it seems that the 77mm X 76mm combo was a pretty crucial part of it from what I've read, but I want to make sure I'm not reading something into it that isn't there.

[Piledriver]

Rings are ~50% of engine friction.
Friction goes up in a linear fashion with bore.
It goes up the square of velocity.(stroke)

I cannot even imagine Vizard arguing otherwise.

I easily beat that fuel economy in a 914 with a bone stock 1700 T4, with well tuned but WAY too large 44IDFs...
For well over 100K miles, and it was absolutely worn out when I started. (The "worn out" was probably the trick)
Same motor swapped into a 71 Super got ~38-40.
Difference? drag. (OTOH I was running 80++ MPH intermittently on a long trip)

Thats a 90x66, for the record.

A bone stock single port in a Ghia can get almost 50MPG, well tuned... It's the low frontal area, not the motor, kids.
The low CR/poor squish hurts things, but that's curable.

FAR more % fuel economy comes from small frontal area, gearing, and excellent tuning that you will EVER see from tweaking engine geometry, or even Nickies.

Invest in a truly tunable EFI/ignition/WBO2 setup, and consider having some stock german jugs Nickasil plated, perhaps look into dry sumping and running some case vac with low tension rings...
Always remember tuning, drag and friction are the process knobs.

The Chevy rod journals have less friction... Ceramic lifters, low tension springs (beehives) all will INDIVIDUALLY probably have more effect than bore vs. stroke, unless you go to extremes.

For a given engine size, a large bore/short stroke engine with the compression ring as high as possible will have the lowest HC readings and effective trapped area, and lowest friction if the pistons/rings are chosen wisely.

Look at Atkinson (or Miller) cycle for a 5-7 % Carnot efficiency boost. Look into cooled EGR throttling.

A small engine, run hard will be most efficient, if that is the ultimate goal. Consider building a twin.

[spectre6000]

Initially I was looking at Atkinson cycle cam tuning, but in order to retain a usable amount of power the stroke ended up needing to be longer than is practical. Further, I lack the skill and machinery to grind a cam appropriately, and in my experience most of the old dudes out there with VW shops aren't sophisticated enough to do anything outside the box.

I want to keep the reliability in check, and while a turbocharger or supercharger (Miller cycle) can certainly be reliable I would feel more comfortable with natural aspiration. I also feel better about my ability to tune a naturally aspirated engine for now.

[Piledriver]

Initially I was looking at Atkinson cycle cam tuning, but in order to retain a usable amount of power the stroke ended up needing to be longer than is practical. Further, I lack the skill and machinery to grind a cam appropriately, and in my experience most of the old dudes out there with VW shops aren't sophisticated enough to do anything outside the box.

I want to keep the reliability in check, and while a turbocharger or supercharger (Miller cycle) can certainly be reliable I would feel more comfortable with natural aspiration. I also feel better about my ability to tune a naturally aspirated engine for now.

Web would be delighted to cut you a 110/86 (or anything you like) combo on almost any lobe center you like for only a nominal cost over a std shelf cam. I had them make me an 86/1600DP cam up awhile back that someday I'll get around to actually using in my Vanagon. (It was actually a regrind on the OG blank, for good reason)

Long intake/short exhaust timing--- How far you go is up to you.

[spectre6000]

I was thinking it would be best to go short on the intake to avoid any potential reversion issues. I have approximately 0 experience with carburetted Atkinson Cycle engines. Assuming the same 76mm stroke, a 10% (36 degree) shorter duration on the intake would put the stroke back at a stock-ish ~69mm on the intake stroke and 76mm on the power stroke. It would be great for cooling too. I also want to use hydraulic lifters (along with full-flow, electronic ignition, etc.) to further modernize the maintenance cycle, but I don't know exactly how that would affect the cam timing.

Have you or anyone else actually successfully run a Type 1 engine with Atkinson Cycle timing? I know Jake Raby was working on something for a customer, but I never saw/heard anything beyond that. I tried to get some knowledge on Atkinson Cycle timing on the Samba and it was like beating my head against a wall...