Modern Trends In Gas Turbine Engine Component Manufacture

Under increasing competitive pressure to improve quality, reduce total throughput times and reduce total life-cycle costs, gas turbine engine component manufacturers are having to rethink their manufacturing strategies. David Drechsler of Huffman Corporation reports.

Various manufacturing strategies have been in and out of favour in recent decades. Major changes in materials, designs or component technologies have necessitated corresponding changes in manufacturing strategies. For example, the introduction of conformal cooling technology and the application of ceramic coatings to airfoils dramatically atered the then traditional manufacturing processes.

In other cases, new manufacturing technologies can affect manufacturing strategy. In recent years, the introduction of super-abrasive grinding technology has changed the thinking in the turbine component manufacturing and repair markets. Traditionally, these markets were monopolised by large conventional abrasive grinding systems, often collectively referred to as 'creep feed grinders'. Multi-axis, super-abrasive grinding technology has introduced new ways of thinking about combining operations, reducing throughput time, lowering cost and increasing quality, especially for complex components like blades, vane segments and shroud segments.

Options

For understandable reasons of safety, design and manufacturing engineers are slow to change proven designs and manufacturing processes on life-critical aero-engine components. But in an increasingly competitive environment, the demand for higher quality and lower total life-cycle cost has driven manufacturers to employ new technologies and methods that were previously not imagined. The costs of acquisition, implementation, operation and maintenance — plus the lead-times, scrap rates and floor space requirements of the old, traditional systems — are no longer acceptable.

Many different manufacturing philosophies have arisen out of the widespread introduction of superabrasive technology. These philosophies are influenced by volume requirements, part geometry, cost or time constraints and individual personal histories of failures and successes. Machines and manufacturing processes run the full spectrum from simplicity to complexity.

At one end of the spectrum are relatively simple, relatively cheap, one-operation machines. These machines are usually arranged in cells where a single operation gets performed on every machine. In an extreme case, for a work piece that requires 10 machining operations, there are 10 machines, each dedicated to a single operation.

The 'simple' strategy described above certainly has its merits. Each individual machine and the process on it can be simple — and consequently less expensive to purchase and own and simpler to learn, operate and maintain.

But there are some negatives associated with this strategy. Firstly, a larger number of machines must be purchased, thereby increasing floor space and possibly overall cost. In practice, there will be queuing time in front of each machine, leading to larger thannecessary WIP and inventories. Also, labour costs associated with handling parts many times over increases. Furthermore, the entire process is vulnerable to any one machine being out of service, and perhaps most of all, every process on every individual machine — clamping, wheel size and condition, coolant application, machine accuracy, and so on — represents an opportunity to introduce process variation.

At the other end of the spectrum are complex grinding systems, some with twelve or more CNC axes, with dual spindles, multiple wheel truing devices, wheel changers, tool changers, truing roll changers, programmable CNC coolant nozzles, on-board or integral CMM inspection processes, pallet changers, robot loaders, and even automated guided vehicles serving the machines.

These systems are sold on the prospects of using high technology to achieve high productivity, low labour costs and even 'lights out' production. The promise of these systems is certainly appealing. In practice, however, in addition to requiring a machine operator, they sometimes require one or two skilled tradesmen plus a full-time manufacturing engineer to keep them running in production. They usually require a massive capital investment and the significant time to get them installed and to get them and keep them in production.

Which Is Best?

This seemingly simple question — Which is best? — is not so easy to answer. 'Best' is subjective. A decision about the best manufacturing strategy and the best machines to carry out that strategy depends on a myriad of factors, including: machine price, lead time, inventory, reliability, productivity, capital, floor space, labour, throughput time, consumable costs, maintainability and many other factors. Usually, the decision involves making trade-offs between many conflicting goals.

Consider the following example. One of the fastest ways to grind a root form on a large turbine blade casting might be with a 100 HP, 3-axis, conventional abrasive, continuous dress creep feed (CDCF) grinder (although new 'hybrid' wheel technology is challenging the material removal rates of CDCF). The material removal rate on such a machine is very high. But a 3-axis machine can't be used to automatically reposition the part to combine other operations, especially more complex radial features.

So the part that required multiple grinds will have to be moved to another machine and set up for other grinding operations — added labour, another fixture, and another opportunity to introduce variation. Sometimes these parts — even mounted in an optimally rigid fixture — cannot withstand the grinding forces of a 100 HP spindle, in which case some of that expensive machine spindle power goes wasted. Further, if the manufacturer wants to run small lots and frequently change setups, then a diamond truing roll must be changed and a new wheel form must be trued into the wheel on each changeover.

So we see that the "best" solution for high material removal rate might not be the best for the overall manufacturing process and for overall productivity. High power spindles are certainly appropriate in many cases. However, power must be considered within the context of other features and desired benefits.

Which Factors Are Most Important?

When asked to complete a forced ranking of 'critical success factors' in their production machines, recent survey respondents ranked reliability and maintainability at the top. Interestingly, reliability outranked things like cost, production rate and even payback time. While not immediately obvious, the survey result is telling. Using the example above, the most powerful grinding machine in the world is effective only when it is grinding; that is, when the wheel is engaged in the work piece. If the machine is not grinding — whether for changing wheels or diamond rolls, truing wheels, or machine down time — the owner is getting zero benefit from that power.

The primary consideration then — regardless of machine, process or manufacturing strategy — is whether the process is robust and reliable. That is, the process must make good parts, every time! The fastest process in the world is no good if the parts it produces are not within specification. Likewise, the most advanced manufacturing process will be crippled if one machine in the cell is plagued with down time.

A Happy Medium?

An alternative between the two extremes is a system that is more versatile than the single-operation machines, but far less complicated and hence more reliable than the 12-axis 'monsters'. Multi-axis super-abrasive machines are versatile, enabling users to reliably and cost-effectively combine multiple operations.

Super-abrasive grinders are especially suitable for combining operations. Users can clamp a part in a single fixture and machine many different features. For complex parts, like shrouded turbine blades, nozzle guide vane segments, and shroud segments, combining operations into one fixture and grinding cycle, where practically and reasonably possible, can produce dramatic improvements in cost and quality. Quality improves since multiple features can be ground from a single set of datum locators. The parts are handled less often, thereby eliminating labour costs and the opportunities to introduce variation.

Super-abrasive wheels are generally smaller than conventional abrasive wheels. Thus it is easier to stack multiple wheels to form a wheel pack. A wheel pack with multiple wheels — each wheel with multiple grinding surfaces — can be used to combine many operations, sometimes 20 or more. CBN technology contributes to a much more stable process. Electroplated CBN wheels never need dressing and vitrified bond CBN wheels require very infrequent dressing. Contrast that with a process using conventional abrasive wheels that requires continuous or frequent dressing, an inherent process variation.

Manufacturers need more agile manufacturing capability, with the ability to quickly change setups, effectively handling smaller batch sizes and improving turn-around time for their customers. Features like quick-change wheel packs with HSK or hydraulic wheel arbours, quickchange coolant nozzle assemblies and palletised fixtures are becoming more readily available. Instead of changing large wheels and having to dress forms into them, well-designed processes enable users to change over from one part to another very quickly, sometimes in well under five minutes.

More and more users are trying to use machining centres for grinding operations and for combining grinding with other machining operations. Of course, most machining centres were not designed to handle grinding forces on high precision components. Nor were they designed to handle high volumes of high-pressure grinding fluids and the corresponding mist and grinding swarf. Machining centres are also ill-equipped to deliver the high-pressure coolant to precise locations required by the grinding process.

Grinder manufacturers have responded by adding wheel changing technology to grinding machines. A wheel changer makes easier work of designing individual wheel forms rather than the more complex wheel packs. To maintain a consistent grinding process, systems with wheel changers must also have the means to change the coolant nozzle assemblies that are usually specific to an individual wheel form or wheel pack.

Wheel changers on high production machining centres or grinding centres have had mixed results. The promised benefits of wheel changers must be tempered by the necessity for a reliable and overall cost-effective process.

A modular machine design is also very beneficial. A machine design that can be configured with three and up to six CNC axes enable a machine builder to tailor a machine depending on the end user's preference. For very high volume dedicated production jobs, a machine can be configured with only those features that are needed for that one type of job. In other cases, users want the flexibility to apply the machine to a variety of jobs — from three-axis blade root grinding to five-axis grinding of multiple surfaces on shroud segments. Finally, programming software has made strong advances in recent years. Some users wish to tightly integrate their legacy CAD designs and CAD/CAM systems to new manufacturing processes, in which case, post-processors can be readily integrated to multi-axis grinders.

Others choose to use menu-based programming systems written specifically for part families like turbine blades, shroud segments, nozzle guide vane segments, or compressor blades with circumferential root forms. These 'family-of-parts' programming systems offer users flexibility in part designs and process parameters and are much easier to learn and use than more complex CAD/CAM systems.

With PC technology integrated into many of today's CNCs, the same programming system on the engineer's desk can be installed for the operator at the machine.

Putting It All Together

Even with the best manufacturing strategies, machines, programming tools and so on, the job is not complete. After all, manufacturers invest to achieve results, not just to acquire machinery. Today, many manufacturers have limited engineering capacity and consequently are more reliant on machine tool builders to deliver proven, robust manufacturing processes along with machinery.

Computerised simulation and development tools help to visualise solutions long before physical hardware is required. This helps to shorten lead times and reduce risk. With modular, multi-axis superabrasive grinders from suppliers with the capacity to deliver well-designed processes, turbine engine component manufacturers can meet today's unprecedented cost, quality and delivery challenges.