Kenneth J. Korane
From all outward appearances, wind turbines are sleek and elegant works of engineering. But inside, they're hell on gears and bearings. In most turbines, the rotors turn a mechanical transmission that drives a small, high-speed generator. In terms of capital investment, this is the most economical and preferred design — but large and unpredictable loads often push gearboxes beyond their limits. Downtime and unreliability are two of the main reasons electricity from wind is expensive.
Low-speed, direct-drive generators that eliminate the gearbox are a more-dependable alternative for turbine designers, but today they can be pricey. Engineers at Timken, Canton, Ohio, may have an even better answer: a rugged and powerful planetary gearbox that can handle extreme loads.
Reliability problems stem from the fact that the wind doesn't blow in a nice, steady stream, says Gerald Fox, chief technologist of mechanical technology at Timken. It's turbulent, changes speed and direction in an instant, and creates loading conditions that play havoc with mechanical systems.
"Turbines have a giant rotor, in some cases as large in diameter as a football field, generating 1 to 2 million lb-ft of torque," explains Fox. "The gearboxes commonly have 75:1 to 100:1 step-up ratios, taking wind energy from the rotors at about 20 rpm up to 1,500 to 1,800 rpm at the generator." The trouble is, says Fox, when wind speed suddenly changes, a relatively small amount of acceleration and angular movement at the gearbox input gets multiplied 100 times at the output — building up massive amounts of torsional windup and strain energy in the gears.
Turbine makers typically use planetary gears to divide torque along three paths and reduce individual loads on each gear. But torsional loads twist gears out of alignment, and slight dimensional variations in gearbox components — including shafts, bearings, gears, and carrier — means planet gears don't equally share the load. Misaligned gears, shock loads, and uneven forces lead to high localized stresses and, eventually, fractures along the gear edges. It also causes bearings to skid rather than roll, smearing and micropitting the raceways and hastening failure.
In essence, gearbox designers face a complex indeterminate problem. Although scientists can carefully measure wind speed and direction, transient dynamic loads are difficult, if not impossible, to measure and accurately account for in the design process. Compounding the problem, notes Fox, wind turbines continue to get bigger — from 1-MW systems a few years ago to 7-MW units in the works today.
One attempted fix has been to install costly, high-precision aerospace gears that reduce manufacturing variations and backlash. But usually, designers try to compensate for uncertainties by increasing safety factors, using larger components, and as many as four rows of cylindrical or spherical roller bearings. These efforts have yielded only marginal improvements while making gearboxes bigger, heavier, and more expensive. And this, in turn, cascades into larger nacelles and beefier towers, further driving up costs.
At Timken, engineers have developed a product called the Integrated Flex-pin Bearing (IFB) to equalize gear loads, eliminate misalignment, and dramatically improve wind-turbine reliability. The IFB consists of a double-cantilever pin supporting the bearing and gear. The cantilevered pin attaches to the carrier wall, and a cantilevered sleeve mounts to the free end of the pin. Gears and bearings mount on the sleeve.
The IFB equalizes loads on planets by anchoring them onto a planetary carrier in a torsionally compliant manner. Instead of fixing the angular position of the planet gears, as is the case with conventional systems, flexible pins deflect circumferentially and independently along the carrier pitch circle — which ultimately equalizes forces on the planets, even while transmitting varying levels of torque, according to Fox.
Simply stated, external forces on the gear make "the pin bend in one direction, the sleeve bend in the opposite direction, and misalignment angle at the gear face remains virtually zero," says Fox.
The first true test of the IFB was in a mechanical transmission from Maag Gear, Winterthur, Switzerland, for a 1.3-MW N60 turbine built by Nordex, Norderstedt, Germany.
The unit is in Scotland's Orkney Islands, a site buffeted by some of the world's most ferocious winds. Nordex had tried various windturbine gearboxes from several major manufacturers but, according to company officials and the wind-farm operators, typical life was far below expectations. The IFBs, on the other hand, successfully equalized loading, reduced internal stresses, and eliminated failures. "They've been operating since April of 2004, so we're now into the fourth year of operation," says Fox. "This accumulated service far exceeds that offered by all previous designs, and the condition of the gears and rollers is like the day they were installed.
"Because shaft, bearings, and gears are completely integrated, there are also significant opportunities for making the units smaller, lighter, and possibly at reduced costs," says Fox. The ready-to-mount IFBs feature a preset bearing clearance, letting gearbox manufacturers shorten assembly time. The new design also offers wind-energy engineers the possibility to upgrade existing turbines with more powerdense transmissions or improve performance and reliability for future projects.
Based on the success of this design, says Fox, many wind turbines struggling with reliability today might benefit from a retrofit with the IFB. The same requirement for more reliability holds for off-highway, aerospace, and other industrial equipment. This has led to the creation of the Timken Flex-drive gearbox.
While still in the concept stage, Flex-drive represents the next generation of planetary gear trains, says Fox. Timken engineers predict Flex-drive systems will increase torque transmission capacity as much as 50% over conventional planetary gearboxes of the same size, letting manufacturers substantially enhance powertrain reliability.
Flex-drive improves torque density using several features. First, it divides torque into more paths to lower the forces and stresses on each gear. It replaces the usual three straddle-mounted planetary idler gears with up to eight IFBs that equalize and reduce loads while minimizing, if not eliminating, misalignment from system deflection and manufacturing tolerance stack-up.
Next, two arrays of three or four IFBs are cantilevered towards each other with a small gap in between. With proper design, torsional windup of the carrier cannot cause gear misalignment. Splitting the IFBs into two independent arrays lets one side index with respect to the other such that when components bend, gear faces remain perfectly aligned. Splitting power on two opposing arrays of tapered roller bearings also means local bending on one mesh doesn't affect and amplify bending on the other.
And in conventional gears, the tooth profile and lead correction can only be optimized for one combination of load and misalignment — generally for severe conditions. This means the full gear face isn't used in normal operation. "Analysis shows that reducing loads and misalignment can virtually eliminate the need for lead correction," says Fox. "Consequently, the Flex-drive can use narrower gears and make better use of effective length. Thus, we can make gearboxes smaller or increase torque density."
Finally, IFBs add circumferential compliancy among adjacent planets and substantially improve load distribution, significantly reducing the safety factor. All this adds up to an opportunity to improve the reliability of existing planetary systems with a "drop-in" retrofit, or create smaller planetary gearing for wind-turbine power trains still on the drawing board.
Fox points out that while the IFB is proven technology, work to date on the Flex-drive is primarily at the analytical level and still needs real-world validation. "The proof is putting it in the customer's equipment, whether a wind generator or an off-highway vehicle. We're currently looking for industry partners," he says.
"The traditional planetarygear world has, to a large extent, reached the point of diminishing returns," he emphasizes. "They can tweak materials and precision to improve performance a little, but gains today are incremental, a few percent at best. Flex-drive is revolutionary, an increase in torque density of 50% is possible, and probably more. And even if we find out that for some applications we can't make every improvement, and the gain is perhaps only 30%, that's still a quantum leap over anything else."
The Timken Co., timken.com