undersquare engine design

Reading about the problems associated with making and using a propeller speed reduction unit on other threads, I got to thinking this:

How about forgetting the PSRU and designing the engine for good low rpm torque with long stroke?

First, some definitions according to wikipedia,
Square engine is: stroke=bore
Oversquare is : stroke smaller than bore
Undersquare is : stroke longer than bore

The book "Aircraft Power Plants" by James L. McKinley says: "It has been found that a square engine provides the proper balance between the dimensions of bore and stroke."

But the small engines I have owned are oversquare. (continental A-65 and Limbach)

Why not more square or undersquare designs for low speed flight such as ultralights or other high drag designs?



Here is a chapter from wikipedia about auto engines:
Undersquare, or long-stroke engine
An engine is described as undersquare or long-stroke if its cylinders have a smaller bore (width, diameter) than its stroke (length of piston travel) - giving a ratio value of less than 1:1.
For example an engine which has 90 millimetres (3.54 in) bore, and 120 millimetres (4.72 in) stroke has a bore/stroke value of:
90 mm / 120 mm = 0.75:1
This can be a negative trait,[citation needed] since a longer stroke usually results in greater friction on the cylinder walls, more stress on the crankshaft,[citation needed] and a smaller bore requires smaller valves which restrict gaseous exchange. An undersquare engine usually has a lower redline than an oversquare one, but it generates more low-end torque.
Engines can be modified with a "stroker" crankshaft, which increases an engines stroke from stock, increasing torque.
Undersquare engines typically are, proportionally, shorter in length, heavier,[citation needed] and taller than equivalent oversquare ones, which is one of the reasons why this type of engine is not generally used.[citation needed]
[edit]Undersquare engine examples
Many British automobile companies used undersquare designs through the 1950s, largely because of a motor tax system that taxed cars by their cylinder bore.[citation needed] Therefore, many of the most famous cars of that era use this design. This includes the Austin A-Series engine, and many Nissan derivatives.
The Chrysler Slant-6 engine, in its most common 225 cubic inch (3.7 litre) version, is a massively undersquare engine, with a 86 millimetres (3.39 in) bore and a 105 millimetres (4.13 in) stroke, producing most of its power right on the peak of its torque curve. The achilles heel of this engine, otherwise known for its exceptional durability, is being over-revved by inexperienced drivers. Red line for a factory engine is under 4,500 revolutions per minute (rpm); red line with aftermarket connecting rods is about 5,500 rpm. On the other hand, a well-maintained Slant-6 can be made to idle as low as 75 rpm (though this is not a recommended speed - neither the alternator nor the oil pump will function adequately).[original research?] In some circles, the Slant-6 is nicknamed "The Stump-Puller" for its diesel engine-like low-speed torque. Appropriate gearing and driving skill is required for performance use.
Willys also used mostly undersquare engines, in fact the L134 and F134 engines, with their fairly small 79.4 millimetres (3.13 in) bore and 111.1 millimetres (4.37 in) stroke, are probably the most undersquare engines ever built (for Jeeps).
The Dodge Power Wagon, among other vehicles, used a straight-six Chrysler Flathead engine of 230 cubic inches (3.8 litre) with a bore of 83 millimetres (3.27 in) and a stroke of 117 millimetres (4.61 in) - yielding a substantially under-square stroke ratio of 0.70.
Undersquare engines tend to be less common than oversquare, but this form of engine is still used in some applications. For example, a modern 8.4 litre Valmet 645 inline-6 tractor diesel engine is a longstroke/undersquare engine, but has an output of over 224 kilowatts (305 PS; 300 bhp) with turbocharger and intercooler.
Numerous Volkswagen Group petrol engines are undersquare.
The popular Mazda Miata also uses an undersquare engine.

Back in the '70s when I was looking for something to power my Maranda project, I looked at a lot of auto engines, trying to find one that wasn't too heavy and had a decent torque at a relatively low RPM. The best I could find in the spec sheets was the Toyota Land Cruiser's inline six. I don't remember the numbers but they were good. I never did find one, so don't know how much it weighs. Most of those sixes and eights were all cast iron, making them extremely heavy. One guy back then hung a Chev 350 on an original design biplane called the Blackbird and had a direct-drive prop on it. I never heard anything more about it, but is was awesomely heavy for the power it produced. There were numerous articles in those days in Sport Aviation about the various conversions guys were trying, but most dropped off the radar and I have no idea whether they were successful or not. I do suspect that any successful setup would tend to multiply and we'd see numerous examples flying.

I never did finish that Maranda. Mortgages and babies tend to dampen such things. I really wanted a deHavilland Gipsy Major for it, since that's what I flew in the Auster VI. It had 145 hp but swung a seven-foot prop that pulled a whole lot better than the 145 hp O-300 Continentals that most 172s had then. And the Auster was the same weight as a 172. The Gipsy, IIRC, had a bore of 4-5/8" and a stroke of 6-1/2". And a redline of about 2600, though I never got it anywhere near that. It was happy cruising at 2200. I'd still love to have a Gipsy; I'd find a way to hang it on the Jodel, and climb straight up.

Dan

Is there a simple expanaiton as to what makes and engine produce peak power at 26-2800 RPM vs 6000? I realize there are many variables but there has to be a common design factor. Is it mainly stroke vs bore?

There are a couple of reasons I can think of for not ending up too undersquare. Adding more stroke makes the engine larger to contain the same bits--larger crakshaft radius moves the pistons farther apart. Generally, that would add some weight. Additionally, increasing the stroke increases mean piston speed (which increases wear and the tendency to detonate), unless the connecting arm length is proportionately lengthened. That makes the engine yet larger. Also, as pointed out for the Slant-6, the undersquare engine tends to have its rpm limited by detonation, unless you reduce compression ratio or just keep rpm low; both of which limit horsepower.

To get that horsepower back, without making the footprint of the engine any larger, would mean adding more displacement by increasing the bore, which brings you back closer to square.

Clean sheet of paper:
How much airflow ( inlet/exh port & valves ) do you need to feed the cyl volume of the engine you decide to build..... this will dictate the smallest cylinder bore size you will need... not hard to work out as @ say 2700 rpm ( max prop RPM example ) & at around 85/90% cylinder filling efficiency at a guess at full power.

Stroke: whatever you need to make the cubes & without going over a mean piston speed of around 2500 feet per minute . ( I can hear you thinking , alright with a 6" stroke @ 2500RPM that is 2500 fpm, just remember the peak piston speed @ approx 1/2 stroke is up around 7800fpm which is very high & can contribute to ring flutter etc which introduces more oil into the combustion area & contributes to detonation.

Rod Length: The old motors some of you mention had very long rods, this builds torque @ low RPM as it 'parks' the piston @ TDC for a longer period when combustion is taking place, this allows the pressure to build higher and as the crank has now moved further past TDC in this time the pressure is applied at a point of greater leverage. Take the SB chev motors as an example 302/327/350 all have 4" bore & 5.7 rod, but strokes of 3.00',3.25",3.50" respectively. The 302 Z28 motors loved to rev, the 327's a bit less & the 350 was more of a cruiser. rod ratio was 1.9/1 on 302, 350 was 1.63/1. Somewhere around 1.7/1 is a ballpark to aim for. Convert your 6" stroke to rod length( x1.7 ) & you get a 10.200" rod length ( about now you can see where the Gypsy rods etc got assembled height from).

Squish: One thing that most of the older A/C engines never seemed to have designed in is a squish band or area in the combustion chamber to help create turbulence as the piston approches TDC, in the automotive Hi-Perf world its the one thing that has helped stave off detonation with todays fuels with lo-lead values, you could say that the dual plug/ignition setup has saved the A/C engines bacon in this dept, but the introduction of very tight 'Squish' clearances in the car industry in recent years is responsible for a large part of the power increases we now have in spite of fuel quality. Put it this way.... If I had to increase compression today I look hard at fitting either a slightly taller piston, longer rod, or decking the block before simply milling the cylinder head. the only downside to this in an Aviation powerplant is if any foreign bodys get in the cylinder with minimal squish clearance you might not get home on the remaining cyls.

Originally Posted by Ivan

There are a couple of reasons I can think of for not ending up too undersquare. Adding more stroke makes the engine larger to contain the same bits--larger crakshaft radius moves the pistons farther apart. Generally, that would add some weight. Additionally, increasing the stroke increases mean piston speed (which increases wear and the tendency to detonate), unless the connecting arm length is proportionately lengthened. That makes the engine yet larger. Also, as pointed out for the Slant-6, the undersquare engine tends to have its rpm limited by detonation, unless you reduce compression ratio or just keep rpm low; both of which limit horsepower.

To get that horsepower back, without making the footprint of the engine any larger, would mean adding more displacement by increasing the bore, which brings you back closer to square.

Detonation is a problem that arises when enough time is allowed for the unburned gases ahead of the flame front to spontaneously combust all at once as the rising pressure increases the temperature of those gases. A low RPM with a high manifold pressure is one of the the surest ways to get detonation; it gives high pressures due to the amount of air and fuel, and the slow piston speed grants time for the unburnt complex fuel molecules to break down into simpler structures that easily autoignite.

In our old North American cars, low RPM was the norm, often cruising at 1800 RPM. Some inline sixes still cruise at that speed. And they will "ping," or detonate, really easily when we step on the throttle without downshifting. Detonation isn't nearly the hassle in today's little high-revving engines; the cylinder bore is small, enabling completion of flame front passage very quickly, and the RPM is high.

Most certified aircraft engines have large bores and low redline RPM. Both of these factors encourage detonation, and so we need higher octane fuels in them unless we have low compression ratios as in many older 80-octane engines. Even then, opening the throttle too rapidly can briefly cause detonation and do a little damage that accumulates over the life of an engine that's abused this way.

The geared engine can avoid detonation just by operating at a higher speed. It can run at much higher manifold pressures because of that RPM. But serious tradeoffs come in the form of a shorter engine life, issues with the reduction as we've been discussing elsewhere, and more weight and complexity and failure points. And cost.

Dan

The Wrights built the first successful airplane engine and it was almost square. This article speculates about the design here: http://www.obrasmechanicos.com/wrighteng.htm Other aviation engines of the time were undersquare. I think maybe the PRSU the Wrights used eliminated the need for long stroke. The Wrights had major problems with the PSRU and this is still a problem today.
Can you believe Charles Taylor built the engine in two months!

I think it might be possible to build the needed larger crankcase by hand, using thin welded steel, it could be lighter than cast. This would allow a long stroke engine to have less weight than an auto conversion with a PSRU. Does that make sense?

Originally Posted by BBerson

Can you believe Charles Taylor built the engine in two months!

Yes, he definitely had no text messages, internet & EAA buddies questioning his every decision, or FAA inspectors wanting everything in triplicate to annoy him! I see they even pulled another engine apart to get ideas, imagine that in todays world, those part manufacturers would be lining up for the royalty cheques before they got off the ground.

Originally Posted by BBerson

I think it might be possible to build the needed larger crankcase by hand, using thin welded steel, it could be lighter than cast. This would allow a long stroke engine to have less weight than an auto conversion with a PSRU. Does that make sense?

Some car makers long ago tried welding up crankcases from steel sheet. Seems to me Franklin tried it, among others. The steel lets the case flex far too much, causing twisting and misalignment of the cam and crankshaft bearings and the cylinders will pull away from it. Cast iron is very rigid and keeps things in proper relationship to each other. Cast aluminum is similarly stiffer than aluminum sheet or plate.

Dan

Originally Posted by Dan Thomas

Some car makers long ago tried welding up crankcases from steel sheet.

Dan

I am looking at the radial engine. This book says radial engines are 40% lighter. A radial engine crankcase may not have a twist problem.
The book has a good deal of info, all of the old engines were oversquare, some 5" bore with 7" stroke.
The Airplane Engine - Google Books