Previous posts have looked at how propellers work and the causes and effects of vibration. This post acts as the third part of a trilogy, showing how this information can be used to select the appropriate shaft and machinery train arrangements for a specific warship. Once a suitable hull form has been derived and the shaft horsepower needed to drive that hull at the specified speed calculated, the question becomes one of transforming that power into thrust with the maximum efficiency.

The first consideration is weight economy. The propeller, shaft and gearing all represent dead weight that is duplicated with multi-shaft arrangements. In addition, the curve of engine output power as compared to size and weight is not linear; two smaller engines together weigh substantially more than a single larger unit of the same output. It may, therefore, seem that an ideal arrangement will involve keeping such duplication, that is the number of shafts, down to a bare minimum. Provided the total power in question is below the maximum that can be absorbed by a single propeller, then a single shaft arrangement would seem to be the most efficient. If the installed power is greater than the maximum that can be absorbed by a propeller, then the most efficient arrangement would be that involving the fewest number of shafts; in most cases two. Another way of saying this is that the most efficient design for shafting is to load the propellers as highly (that is, to put as much power through) as possible.

For merchant ships this is indeed the case. Merchant ships are designed for economy of construction and use, not for the most efficient use of high power settings. In their case, the economic advantages of a single shaft outweigh any disadvantages from the layout. What this really proves is that merchant ship practice does not carry over into warship design. It is not possible to make arguments for a given configuration for a warship by quoting merchant ship practice. The demands of the two are so different that a comparison between, for example, a liner and a battleship are essentially meaningless.

A single centerline shaft turns out to be a very poor choice for a heavy warship. One problem that's immediately obvious is the dangers of damage or mechanical failure. If that shaft is damaged by, for example, mine or torpedo strikes, bearing failure or any of the other hazards of being a combatant warship, the ship is helpless until the damage is repaired. Experience has shown that accidents and mechanical failure are more of a problem than combat damage but the principle holds; a single shaft exposes the ship to appreciable risk. The same applies to machinery; if a ship is powered by a single engine, then she is held hostage to the reliability of that engine.

There are, however, more serious problems with a single shaft. One is that a ship so equipped cannot use her engines for steering. Below about 10 knots, a ship's rudders become ineffective. In this environment, a single shaft ship is uncontrollable and needs to have tug assistance for docking or other maneuvering requirements. This can be partially cured by using twin rudders that flank the single screw and direct the race from that screw. This reduces the minimum effective speed for rudder control but does not cure the problem completely. Multi-shaft ships can use differential power from their engines to bring about steering control, intrinsically a much safer and more satisfactory situation.

Another mechanical problem with single shaft layouts is the fact that the shaft has to be along the centerline of the ship, for most of its length above the keel. The problem here is that the keel is also the foundation for and primary support of the heavy gun turrets. This point became critical in the era just before WW1 when the weight of gun turrets increased rapidly as gun caliber moved inexorably upwards. One effect of this was to make wing turrets (which obviously could not use the keel as their primary support) less viable. Unless an all-forward armament solution is adopted, we have an immediate design conflict that is extremely difficult to resolve. The only way heavy gun turrets and a centerline shaft can be accommodated is to provide heavy carry-through structures that distribute the weight of the turrets (similar structures were used for wing turrets). These structures are heavy enough to completely eliminate any weight efficiency gains resulting from the use of a single shaft. To make matters worse, they act as a transfer medium by which shock and explosion damage can be carried through from the sides of the ship to the centerline, offsetting the added protection apparently afforded by burying a shaft deep in the hull structure (reports in Dubious and Ghastly of the damage suffered to the center shaft of Scharnhorst illustrate this).

Another serious problem with centerline shafts is vibration. The torsional vibration within the shaft itself cannot be cancelled and will be a constant factor afflicting the ship. The screw itself is operating in the turbulent wake of the hull structure, causing pulses of vibration as the blades hit the turbulence. To make matters worse, this screw is directly under the ship's keel so the vibration pulses strike the centerline of the hull and are immediately transmitted through keel and distributed throughout the ship's structure. Other vibration pulses, travelling down the centerline shaft pass through all the structural nodes of the ship, spreading them throughout the hull structure. The heavy carry-though members provide excellent vibration transmission paths and add to the problems. In merchant ships, these problems are not that serious; merchant ships do not usually use the power settings and use profiles that make vibration a serious concern although the single shaft on most merchant ships does give a characteristic and unpleasant thumping in the aft sections.

For warships, these vibration problems are of grave concern. Yet, they are not the prime problem for single shaft layouts. The main killer for these designs is that centerline propellers are grossly inefficient under the conditions prevailing in warships. For true efficiency, propellers have to act in smooth water yet a centerline prop is, by definition, surrounded by the turbulent wake from a ship's hull. The effect on the propeller's efficiency is devastating. Investigations quoted in the earlier posts have shown that between 15 and 45 percent of the energy supplied to a centerline propeller at high power loadings is lost in inducing vibration within that propeller and its surroundings. In contrast, the equivalent figures for wing propellers are between one and four percent. Related to this is a more fundamental point. As we have already seen, propellers work best when turning slowly, that is when lightly loaded. Their efficiency drops rapidly as loading increases. Therefore, two slow-turning, lightly loaded propellers use power more efficiently than a single faster-turning propeller of equal size.

The next question is, given the propulsive advantages of adopting twin screws over single shaft layouts for surface ships, does the configuration bring any specific disadvantages with it? The first is that a twin-shaft solution requires additional space in the rear end of the ship in order to ensure that there is enough space between the screws to prevent unfavorable interactions. The layout also means that the screws will be closer to the sides of the ship, a feature that will give problems in designing the torpedo protection system in this area. The vulnerability of the shafts is enhanced by the fact that the hulls lines aft mean that those shafts run outside the hull for a substantial proportion of their length. Although this is beneficial in that it reduces the level of vibration transmitted to the hull, it does add to vulnerability compared with a centerline shaft buried within the hull structure.

A factor related to damage control is the engine room design itself. A twin-screw ship will usually require a larger engine room (to accommodate the two engines required for its shafts) than that needed for a single shaft ship. If breached and subject to flooding, this larger volume represents a greater proportional danger to the ship. However, this point is often overstated. The most immediate danger resulting from damage to an engine room is not flooding but loss of power from the generators invariably co-located with the main engines. The time taken to switch from this power to emergency back-up generators can be critical. In this context the dimensions of the engine room are of little consequence. It could be argued that a twin-shaft layout actually has some advantages since it allows the installation of a centerline bulkhead that could restrict damage to one engine room and preserve the other, and its generators, from flooding. Centerline bulkheads are very controversial since they can also cause asymmetric flooding and foster capsizing. Japanese designers liked them; US designers abhorred them. The point is that the choice of shaft arrangements does not compel a decision one way or the other and this is a matter best left to the Instructions to Designers.

On balance, these considerations add up to a conclusion that the gains from using a twin-shaft layout greatly outweigh the weight inefficiency of doubling up on shaft, propeller and gearing. The only time when a single shaft is acceptable is where we have a mobilization design and the primary requirement is to keep the number of engineering bottleneck components to a minimum, the recent US FF and FFG classes being good examples. Here, the driving requirement was to keep the gear cutting to a minimum since this has historically been one of the main bottlenecks in ship production. Another case is where the ship's lines aft are so fine that doubling up on shafting is not practical. In this context, modern submarines, almost invariably single shaft designs, are a very special case due to their highly specialized hull lines.

The next question is; since using twin shafts as opposed to a single centerline shaft shows such great advantages; what happens if we double again and go to a four-shaft solution? Do we see further gains or does the law of diminishing returns apply?

The weight economy arguments against going from twin to a quadruple shaft layout are effectively a repeat of those against going from single to twins. The gearing, shaft and propeller are effectively deadweight while four small turbines weigh more than two larger ones of the same aggregate output. Much of the pro-argument follows along the same lines as well; assuming the screws are the same size, they can be much less heavily loaded and, therefore, operate more efficiently.

The big negative on quadruple screws is that they require a broad aft section; four shafts simply cannot be fitted into a finely tapered stern section without serious design problems. The screws have to be spaced out to prevent unfavorable interactions. In reality this means that the choice is often not between two and four screws of the same size but between two large and four smaller screws and the question now becomes one of the relative propulsive efficiencies for that particular design. However, there is one factor here that is interesting; with proper design, it is possible to arrange quadruple screws so that there is a small but appreciable benefit in propulsive efficiency from the races of the screws and their interaction with each other and the hull lines. This benefit is usually between two and four percent on propulsive efficiency; not a great amount but one worth having.

A major plus for quadruple screws is that the arrangement allows for effective active cancellation of torsional vibration in the shafts. In effect, the shafts on each side can be designed to cancel their torsional vibration and then the pairs on the opposed sides balanced to smooth out what's left.

A big advantage of quadruple screws is damage control. Engine steering in the event of rudder or stern damage is greatly eased. Internally, having four engines opens many possibilities with regard to dividing up the engine rooms and makes plausible the idea of controlling lists from engine room damage by counterflooding opposing areas while maintaining the watertight integrity of others. The important thing is that quadruple screw layouts do not, of themselves, force the designers to any particular solution for compartmentation of the engine room spaces; whether to install a centerline bulkhead remains an option of the designer as determined by the appropriate ItD. Put another way, the use of four smaller engines opens up options not available with different configurations.

So, if four screws is a desirable (but space consuming ) solution, can we gain anything by going a step further and installing six screws? Here, the evidence seems to be that we've hit a point of declining return and that the combination of a very wide hull forced by this arrangement, the dead weight of all the additional shafting etc and possible interactions between the screws means that the penalties outweigh the benefits. Or seem to; the only example I can find of a ship with six screws is the old Russian Popovka class river defense ships. These failed quite badly.

Heading the other way, if, on a given power output, four screws is efficient but space and weight consuming and two screws uses weight more effectively but shows less propulsive efficiency, would a triple screw layout offer a good compromise? A preliminary examination of the figures suggests that it might; a comparison of machinery weight per SHP output between ships using triple and quadruple shaft layouts does show an appreciable advantage to the former. However, as we have seen, this is not the whole story.

Firstly, we are comparing numbers between two ships from two different countries. This is always dangerous since no two countries measure such statistics the same way. There is a strong probability that one set of figures contains components that the others do not. Even if this is not the case, weight economy is only one part of the equation. Propulsive efficiency and vibration are of greater significance as is the effect of the arrangement on the ship as a whole.

Here, triple shafts combine all the worst problems of a single-shaft layout and a twin shaft system. About the only advantage of the triple shaft layout is that it eliminates the vulnerability of the single shaft layout to mechanical damage or accident. The design hydrodynamics is such that the effects of the centerline screw actual degrade the efficiency of the wing propellers. In his memoirs, Admiral Scheer made the following comments on his (triple shaft) battleships:

"The advantage of having three engines, as had each of these ships, was proved by the fact that two engines alone were able to keep up steam almost at full speed; at the same time, very faulty construction in the position of the engines was apparent, which unfortunately could not be rectified owing to limited space' Thus it happened that when a condenser went wrong it was impossible to conduct the steam from the engine with which it was connected to one of the other two condensers, and thus keep the engine itself working. It was an uncomfortable feeling to know that this weakness existed in the strongest unit at the disposal of the Fleet, and how easily a bad accident might result in leakages in two different condensers and thus incapacitate one vessel in the group."

This excerpt has two valuable insights. One is the confirmation that the German ships could maintain speed using their wing shafts only; an indication of the inefficiency and redundancy of the center shaft. The other is the suggestion that the center shaft itself was seen as being a reserve against mechanical failure and/or battle damage. The comments about condenser problems are also interesting but by no means unique. "Condenseritis" was a well-known and pervasive problem with all ships in WW1 and its prevalence in the German fleet should not be seen as unusual.

Triple shafts come into their own where there is a requirement for high output power in a hull with extremely fine lines aft. This was the motivation behind the use of the configuration on the Ark Royal and Illustrious class carriers (the combination of treaty limits restricting the length of the armored box, the need for beam and high installed power all conspired to give the designers heart failure). When the treaty limits were lifted, the British redesigned their carriers (Indefatigable and Implacable) with a conventional four shaft layout.

The final question is, would a five-shaft layout show any particular advantages or limitations? On purely theoretical grounds, it's difficult to see why this would be adopted; it would simply share the problems of four-shaft and single shaft layouts. I don't think a five-shaft design has ever been seriously considered - if someone does know of such a design, could they please point me to it???

Single shaft

Advantages

Good weight economy in power train components, shaft buried in hull for protection, economy in war-critical production bottleneck items.

Disadvantages

Inefficient power utilization, high noise and vibration levels, no redundancy against mechanical or combat damage. No engine steering capability. Severe design problems with regard to other parts of ship.

Double shaft

Advantages

Relatively efficient with reduced noise and vibration, allows engine steering, and provides redundancy against damage.

Disadvantages

Requires wider section aft. Preferred design for smaller warships - say up to light cruiser size.

Triple shaft

Advantages

Allows increased power through narrow stern section.

Disadvantages

Inefficient power utilization, high noise and vibration levels. Severe design problems with regard to interaction of power train configuration with other parts of ship.

Quadruple shaft

Advantages

Very efficient due to favorable prop interactions with reduced noise and vibration, allows engine steering, and provides redundancy against damage. Also, allows flexible subdivision of machinery spaces.

Disadvantages

Requires wider section aft. Preferred design for larger warships - from heavy cruiser to battleship size..

9 July 1999

Updated.