Balance - Flat-six vs. Flat-four

[Alfred1]

There was a discussion recently on the other board comparing the 914-4 and 914-6 and there was the usual debate about which engine is better so I thought I’d scan some pages from my "Design and Tuning of Competition Engines" book.

Flat-twins are in complete primary (page 130) and secondary balance (page 135) and since the flat-six is "in effect three horizontally-opposed twins side by side" it is also inherently balanced (pages 143, 144). On the other hand, the flat-four is in primary balance but the secondary forces of the motor tend to set up a “rocking couple” (page 137).

The six also receives more power pulses per crankshaft rotation than the four for smoother running.

Pages 130, 131

Pages 132, 133

Pages 134, 135

Pages 136, 137

Pages 138, 139

Pages 140, 141

Pages 142, 143

Pages 144, 145

[lapuwali]

Yes, flat sixes are well known to be one of the best balanced engines possible. However, the other difficulties with the design (mostly packaging) make it unpopular. The V6 loses some of the inherent advantages, but is still pretty close, and enjoys a layout that's much easier to put in most cars, which is why the V6 is so popular.

Inline sixes are also nicely balanced, and used to be reasonably popular, but now only BMW is using that design much, and mostly due to packaging concerns. Engines are expensive enough to design that manufacturers want to be able to a given design in a wide array of cars, and the inline 6 is really only practical in a front-engine/rear-drive platform.

To put inherent balance on the scale of importance, however, is to note that Wankel rotaries are very nicely balanced, and turbines inherently in perfect balance. However, there are enough problems with both of these designs that they're essentially ignored today (yes, except for one car produced by Mazda). The inline 4, however, is far and away the most popular design today, yet it has inherent balance that's substantially lower than a V4 or flat 4. Packaging and simplicity (only one head casting, much simpler exhaust design, can be used transversely or longitudnally) simply win out over balance issues. This is so much that case that inline 4s larger than 2.0L are so badly balanced that most such designs have some sort of balance shaft (or two) to cancel out the vibration. Even high performance cars (later 944/968) will resort to power-sapping balance shafts rather than use an inline or vee 6 of equal capacity.

[Will98D]

my SAE engine design book is still packed away so it's hard to document right now, but I believe the flat twin has nowhere near the smoothness of a number of other engines because in addition to the primary and secondary forces of rotating and reciprocating masses you have to remember the uneven firing pulses, which make it not much better than a single. the inline 6 is among the best, the 90 degree V8, V12 and V16 good as well. V6 engines leave a lot to be desired. In addition to BMW using I-6 engines for smoothness and favorable torque characteristics, GM also employs an inline 4.2 liter I-6 (275hp) in their mid-size SUVs. I believe you'll also find an inline 4 is far better than a V4.

In addition to the problem of the length of the I-6 packaging, you also have a long crankshaft that has to deal with the challenges of flex and torsion.

[lapuwali]

One has to distinguish between a "180 degree vee" and a "boxer" engine, as well. Both are "flat" in that the cylinders are horizontally opposed, but one has both pistons going right/left at the same time, whereas the boxer design has the pistons coming in together, and going out together, thus requiring separate crank throws for each cylinder, unlike the common vee configuration of sharing a throw for each pair of opposing cylinders. They have substantial differences in balance. The 180d Vee does, as Will states, have terrible balance characteristics. The boxer, however, has very nice primary balance, decent secondary, but a rocking couple forced by the separate crank throws, which tends to make the crank ends try to describe a circle.

I don't know of any mainstream designs that used the 180d Vee in favor of the boxer layout. Ferrari used a wide angle Vee in the early 60s on some racing engines, but the VW, Porsche, and Subaru flat fours and sixes are all boxer designs, as are the BMW (bike) and Citroen (2CV) flat twins.

A 60-90d V6 has wonderful balance characteristics, so perhaps Will is misremembering his SAE book. Not as wonderful as a boxer 6, but still very good, with better packaging. The Alfa 60d V6 is turbine smooth.

For those of you who haven't read any engine and chassis design books, reduced vibration is not just a matter of comfort. Vibration can easily destroy an engine, and too much imbalance significantly limits peak revs available (and thus peak power). An inherently well-balanced engine requires much less reinforcement in the block and less substantial bearing saddles to support the loads, which means the engine can be smaller and lighter for a given power output, or can produce higher revs (and power) for a given weight and size. If the engine itself is lighter and vibrates less, this also means the engine mounts and the engine cradle can be lighter, which creates a nice cascading effect with making everything else lighter. This improves performance in a wide variety of ways, from straight line acceleration to cornering power to fuel economy.

[Alfred1]

It's still not clear to me what secondary forces are out of balance with the flat-four. For every crankshaft throw there is one that points in the opposite direction and so the acceleration of those two pistons would always be equal in magnitude but opposite in direction and those forces should cancel each other out. The same would hold true for the other pair of crankshaft throws. I think that fact indicates the flat-four is in primary balance but what exactly is the secondary "unbalance"? Is it that the crank throws for one side of the engine point in opposite directions but the forces associated with those two pistons are not always equal but opposite in direction?

[Alfred1]

I did some calculations to try to figure out exactly why secondary forces set up a “rocking couple” in flat-fours.

The calculations I did show that the crankshaft throws that are 180 degrees apart and with their connecting rod and piston assemblies on opposite sides of the engine have accelerations that are always equal in magnitude but opposite in direction to one another so that the forces of their connecting rods and pistons should always cancel each other out. For the crankshaft throws that are 180 degrees apart but have their connecting rod and piston assemblies on the same side of the motor, I calculate that those accelerations (and therefore their forces) and not always equal and opposite. If I did my calculations right, the sum of these unbalanced forces is periodic and peaks every one-quarter turn of the crankshaft. Here’s the graph.

[Will98D]

OK, finally dug up the book (Advanced Engine Technology by Heinz Heisler, SAE Int'l Publications, copyright 1995; Library of Congress Nr 95 74841). Anyhow, the only engines listed in perfect balance without the need for counterweights or external (flywheel & front pulley) are the inline-6, flat-6, and 60-degree V-12. In the case of the 60deg V-6 and 90deg V-8 with 2-plane crankshaft they can balance out 100% with the above weights. Some never balance out 100%. With the flat-4, it is indeed a couple created by the secondary forces that causes the imbalance. So, a difficult purist would say only those first three are naturally balanced, but there are more you can get 100% with added effort.

Interestingly, the guy who wrote it is in the UK, and he doesn't even address the balance of the 90deg V-6 although he does mention the engine later when discussing offset crank throws. Must be they figure it's low-brow to just chop two cylinders off of a V-8.

[Alfred1]

Heisler's book is next on my list. I decided to buy John Heywood's book first but unfortunately it doesn't cover engine balance but I realized that another book I just bought used for a few dollars covers engine balance in detail. Btw, consider the reputations of the car manufacturers that use a lot of the engines that are inherently balanced - Porsche, BMW and Ferrari.

[BigD9146gt]

So Alfred, what does the flat 6 and V12 graphs look like compared to the flat 4?

[Alfred1]

Originally posted by BigD9146gt
So Alfred, what does the flat 6 and V12 graphs look like compared to the flat 4?

Now that I've read through some of my book on the kinematics and dynamics of machinery I see that analyzing the balance of reciprocating engines is much more involved than what I was doing with my calculations. The book devotes at least two chapters to analyzing reciprocating-engine balance and the only calculations that I did that were valid were the ones where I calculated the accelerations of the connecting rod ends. Those accelerations would probably only be representative of the actual forces on the crankshaft if the crankpins and connecting rods had no mass. Here are 14 of approximately 100 pages on engine balance from the book. Maybe later I can plug some specs into these equations and graph them to see the net forces and torques for the different motors.

Engine Balance 619

Engine Balance 620

Engine Balance 621

Engine Balance 622

Engine Balance 623

Engine Balance 624

Engine Balance 625

Engine Balance 626

Engine Balance 627

Engine Balance 628

Engine Balance 629

Engine Balance 630

Engine Balance 631

Engine Balance 632