Published on: 


Gearboxes are widely used to reduce or increase the speed of prime movers to match them to driven machinery. Especially in critical operations where redundant equipment is not installed, the gearbox is as important to reliability as any other link in the power train.Yet many engineers tend to be more interested in the gas turbines, electric motors, compressors, and pumps than in the gearbox.

As the power and power density of rotating machines continue to increase, however, turbo gearboxes are approaching their limits in terms of safe and reliable operation. Some radical re-thinking in gearbox design is called for if future machinery trains are not to be compromised by gear unit capacity.

Conventional limits

Traditional gear designs for turbomachinery fall into two types: Parallel shaft and epicyclic.

Parallel shaft gearboxes have two shafts, each carrying a single gear. They are available in power ratings up to 140 MW for small gear ratios <6 when installed, for instance, between a gas turbine and a generator. As the name suggests, the input and output shafts do not share the same axis.

Epicyclic gearboxes split the power between several “planet” gears which move around a central “sun” gearwhile alsomeshingwith a surrounding “ring” gear. Epicyclic gears are compact and provide co-axial input and output, but are limited to a power rating of 45MWin the demand for a gear ratio >6.

To understand the limiting factors in gear design we need to translate external operating characteristics — power ratings and speeds — into gear-specific design parameters. The main factors controlling power and speed limits are:

• Pitchline velocity (PLV) — the linear velocity of the gear teeth

• Elastic deflection produced by torques and bending moments on all parts of the gearbox, especially as it applies to the pinion (the smaller of the two gears in a parallel shaft gearbox)

• Desired input and output speeds

• Operating limits of the bearings (loads and journal velocities)

• Factor of safety chosen for the application.

The job of the gear designer is to find the best balance between these limiting factors, which sometimes conflict. For instance, taking advantage of the increased speed capacity of a high-performance bearing implies an increase in the PLV, which in turn raises the circumferential forces within the gears and the tendency for oil flow to be interrupted. In a parallel shaft gearbox the solution is to increase the center line distance (and hence the overall gear size) to create a larger lever effect.

Currently the maximum achievable PLV is 200 m/s and the maximum journal velocity is 100 m/s. The maximum bearing load is 3.45 MPa if we refer toAPI 613 (paragraph, the gear standard applicable to high-speed gearboxes in the oil and gas industry. These limits mean that in some cases the gear unit becomes the determining link in the drive train. Turbocompressors, for instance, are currently restricted to approximately 35 MW by their gearboxes, depending to some extent on the input and output speeds required.

Incremental design improvements, such as the use of high-grade steels to withstand high PLVs or an improved bearing design for higher journal velocities, boost the capabilities of conventional gear units. The resulting improvements are relatively small.Also, it is a bad idea to be using such prototype gear units running in unproven areas of the speedpower chart.


Instead of waiting for incremental improvements inmaterials science, one solution is an alternative gear design with a split power path. Such an approach overcomes the limiting parameters controlling the teeth and bearings of conventional gear types.And it removes the bottleneck imposed by current gear designs on the power limits of turbomachinery chains.

Splitting the power path to create a power distribution gearbox (PDG) allows the speed limit to be increased and eases the bearing problem. The limiting factor in the design becomes the toothing: The speeds at which the teeth mesh and the heat generated there,mainly through compression of amixture of oil and air. PDGs can currently be designed for power ratings up to 170 MW, output speeds up to 100,000 RPM and gear ratios up to 10.

PDG design and function

The PDG combines the robust construction of parallel shaft gears with the load-splitting ability of epicyclic gears (Figure 1). The result is a gear unit that offers high power ratings, speeds and speed ratios while remaining well within critical technical limits and safety margins.

The PDG uses a planetary arrangement to divide the incoming shaft power, compound gears to further increase the peripheral velocity and a second planetary arrangement to drive the output shaft. The result is a gear unit that can transmit more power at higher speeds than is possible with a conventional parallel shaft or epicyclic gearbox.

The PDG contains three or more planet gears arrayed around a central sun pinion. Each planet gear is a compound design whose second set of teethmesheswith a second sun pinion on a separate shaft. This arrangement acts like a two-stage parallel shaft gear. However, distributing power between planet gears reduces tooth loads. Moreover, no bearings are required on either the low-speed or the high-speed side, and the three planet gears are supported by six hydrodynamic bearings instead of the four used in parallel shaft gear units.

Additional advantages include the fact that the input and output shafts are coaxial. This enables shorter andmore compact drive trains (Figure 2). As the two sun gears are supported by the planets, the main input and output shafts need no bearings. Integral double diaphragm couplings on the input and output shafts accommodate axial, radial, and angular offsets to ensure optimal power transmission and rotordynamic performance. The short axial length of these couplings reduces the overall distance between shaft ends (DBSE) between the driver and the driven machine.

Critical applications

An example of a typical gas-turbine-driven compressor shows how the PDG performs compares to a parallel shaft gear unit in applications requiring high power density. The gas turbine speed is assumed to be 5,000 rpm and the compressor speed 12,500 rpm. In this case the parallel shaft gear reaches 190m/s, nearly themaximumsafe PLV of 200 m/s. For the PDG, the PLV curve rises more slowly, and at 26 MWits PLV is 30% lower (55 m/s lower in absolute terms) than the parallel shaft unit.

Even at power ratings up to 65MWand beyond, the PDG still has a PLV below 180 m/s (Figure 3). Performance is improved because of the smaller diameter of the planet gears in the PDG that increases the safety margin for centrifugal forces, which are proportional to the square of the velocity divided by the diameter.

The curves relating bearing journal velocity to power transmitted are similar to those for PLV against power (Figure 4). The parallel shaft gear cannot handle more than 26 MW before the journal velocity reaches its safe limit of around 100 m/s.

At this power rating, the PDG bearings are running at 50m/s.At 65MW, the journal velocity is nomore than 65m/s. Lower journal velocities arise because the pins of the planet gears in the PDG are smaller in diameter than the radial bearings which support the pinion in the parallel shaft gear unit.

Along with pitch line and journal velocities, bearing load is the thirdmain factor limiting gearbox performance. Even at high powers, increasing the diameter of the planet pins keeps the PDG bearing loads below the maximum of 3.45 MPa set by API 613. Bearing loads are also reduced because the PDG has six bearings (two for each planet gear shaft) compared to the four supporting the gear set of the parallel shaft gear.

Also in terms of efficiency, the PDG rivals the parallel shaft design. As the parallel shaft gear approaches its limit of 26 MW, losses increase significantly, and at 26 MW the efficiency of the PDG is 0.7% higher. Oil consumption of PDG and parallel shaft gear is identical.

Splitting the power path, then, gives the PDG more headroom in PLVs, journal velocities and bearing loads. Lower stresses and temperatures improve reliability while efficiency remains high. However the higher complexity of a PDG means that some design aspects have to be considered.

To ensure sufficient tooth lubrication, for instance, the planet gears require internal oilways as well as external flush lubrication. Within the gearbox, the relatively high ventilation, which is at least proportional to the square of the velocity, can disrupt external lubrication and prevent the sun gears from getting enough oil. Special oil baffles are needed to tackle this problem, and there is room for further improvement.

As a result, the PDG is best suited to applications requiring high power density, where traditional gear designs are approaching the limits of feasibility. In particular, this gear design brings opportunities to debottleneck drive trains in which the gearbox has previously been the limiting factor. Gas-turbine- driven compressor trains, for instance, can nowuse larger compressors, and gas turbine models, which up to now have been used mainly for power generation, can find new applications as compressor drivers.


Tome Stanic is Head of Business Development at Voith Turbo BHS Getriebe, Germany. The company has specialized in high-speed gearbox engineering and manufacturing for more than 80 years. Contact: tome.stanic@voith.com