The waterboxer is a horizontally opposed, 4-cylinder, 4-cycle engine just like its older Type 1 and Type 4 brothers, except it is water-cooled instead of air-cooled. What is interesting about the VW design is that the crankshaft has only (3) main journals. If you were to look down at one of these engines while it was running, and you had X-ray vision, you would see the forward two pistons (#1 forward right and #3 forward left) and the two rearward pistons (#2 rear right and #4 rear left) reciprocating back and forth away and toward each other, but slightly off-set from each other front to back. It's actually like two 2-cylinder engines joined by one crankshaft, or like two BMW motorcycle boxer engines bolted together.
Like any flat-four, 4-cycle engines - the front two and rear two opposing pistons get to top dead center (TDC) at the same time. One of them is just about to fire (compression/power stroke) and the other is exhausting (exhaust/intake stroke, or "overlap"). When the two front pistons (1 and 3) are at TDC, the other two (rear) pistons (2 and 4) are at bottom dead center (BDC). Every time a spark plug ignites and an explosion takes place above the piston, the force is transmitted down the connecting rod to the crankshaft rod journal causing a "racking" or twisting force around the vertical access, over and over, again and again, billions of times over the life of the crankshaft. Since there are only 3 main bearings, all of these twisting forces pass right through the center main bearing. It is for this reason that the crankshaft all VW flat-four engines would eventually crack right through the center main bearing journal, usually at about a 45-degree angle. Interestingly, since the break occurred at an angle, the engine would often continue to run for some time even with a crankshaft broken clean in half! If it did not break there, it would at the #1 journal where the flywheel bolts on because that is the next weakest point. The Type 4 engine design improved the #1 journal strength problem by making that journal significantly larger than the earlier Type 1 engines—switching from one flywheel bolt+4 dowels to 5 bolts—reducing the failures at that end considerably. The center main journal failure persisted, however.
The standard 2.1 liter waterboxer crankshaft is not known to fail at the center main journal. As a matter of fact, we have NEVER seen a water boxer crankshaft break at the center main journal-EVER. It is a superb design, way more sophisticated than any crankshaft used in the two earlier VW horizontally-opposed engines, the Type 1 and Type 4. The waterboxer crankshaft design overcomes the twisting problem by implementing two massive sections, one between each pair of opposing rod journals—that are not present in either earlier design. These two hunks of mass act kinematically to absorb the momentary impact of the ignition process, thereby alleviating the stresses to the center main bearing. The waterboxer crankshaft also employs the identical #1 main journal design that VW figured out worked well on the Type 4. This is just one example of how VW drew on its many decades of knowledge, which culminated in the waterboxer design.
We go one step further on our 2300cc and larger displacement engines by adding counter-weights. Counterweights are not added to horizontally opposed engines to make them more "balanced" or "smoother." Horizontally opposed engines are already harmonically balanced due simply to their physical layout. The counterweights are added to reduce dynamic stresses, so the engine lasts longer.
That is the reason we counterweight the crankshafts in our larger engines: to decrease dynamic stresses and increase longevity—and since we there is already welding happening to increase the stroke. The crankshaft we use in our 2.3 and larger engines is increased from the standard 76mm. In the case of the 2.3, we increase it by 3.5mm to 79.5mm; the 2450 and 2700 crankshafts are increased 8.5mm to 84.5mm stroke. Since we don't know how much margin there is in the original design (that is, how much more stress it can handle and last) we decided to play it safe. We choose to add counterweights to offset the greater stresses created by the longer stroke just for good measure. And since the crankshafts are getting modified anyway to increase stroke, the cost to add the counterweights is minimal, making the process cost-effective. We could save about $100 per engine, eliminate the counterweights, and it would work fine—it just would not last as long.
But alas, nothing lasts forever: crankshafts can and do eventually crack and fail. All the crankshafts we use on all of our engines are used parts to begin with. Welding on them does indeed increase the likelihood that any small defects can turn into cracks. Every single crankshaft we have welded, stoked and counter-weighted is crack tested using a magna-flux machine before it is used in one of our engines or sold to anyone. It is the very last step, as a matter of fact—which means all the time and money has already been spent on the crankshaft—down to the final polish. Nevertheless, we end up scrapping about 1 in 10 crankshafts due to flaws detected that have any chance of turning into a crack. It is a painful but necessary part of building reliable engines.
You might ask why we just don’t use new crankshaft instead. Well, we did that for a period of time. We got one of the best brand names in the crankshaft industry to make us brand-new-from-scratch crankshafts. The problem we ran into was that, in the quantities we needed, it was not possible to duplicate the super strength and quality of the original VW, forged crankshaft -- which was made in the hundreds of thousands. The cranks we had made were machined from solid billet which A) was not quite as strong and B) was simply not made to the same standards as the original, German-made crankshaft. All in all, the failure rate was 10x greater than that for a welded OEM crankshaft.
Still, no crankshaft lasts forever, no matter how it is built or how carefully it is inspected. We have ongoing crankshaft failures—a few dozen as of September 2022. The failure is almost always at the flywheel end where VW improved but did not completely eliminate the failure issue at the #1 journal. A “few dozen” might sound like a lot. Indeed, it represents an enormous amount of cost and heartache. But after building some 3000+ engines and counting since the mid-'90s, that total number of crankshaft failures is actually a tiny percentage. We’ll take it!