Printer Friendly

Castings Propel Space Shuttle Improvements.

Nasa has turned to metal castings as the solutions to part and cost reduction on its Space Shuttle Discovery's main engine.

The world was captivated In late October 1998 when famed astronaut John Glenn and six others rocketed into space aboard the Space Shuttle Discovery, arguably America's most complex, sophisticated and high-tech machine ever created. As the Discovery smoothly launched and disappeared into the heavens, TV announcers informed the viewing masses of the increased performance, safety and reliability of the vessel. Yet what the masses likely didn't realize is that the improved design of several high-tech key engine components was the result of metalcasting. Because of 17 cast metal components in the Space Shuttle Main Engine (SSME), NASA capitalized on a better performing, safer engine, as well as a substantial reduction in total costs.

Now in its 40th year, NASA, like the rest of the world, has found itself under scrutiny for improved cost and performance demands. To meet this challenge, NASA and Pratt & Whitney, its contractor for the main engine turbopump, have implemented programs over the past several years that convert costly welded, assembled and forged components into one-piece castings that are more economical, safer and reliable. This article examines the role that castings play in the SSME.

Application in Action

When a space shuttle is lifted off the launch pad, it does so with the help of three reusable, high-performance rocket engines that provide 400,000 lb of thrust. Each of the powerful main engines is 14 ft long, weighs 7000 lb and is 7.5 ft in diameter. The liquid hydrogen/liquid oxygen engines fire for about 8.5 mm during liftoff and ascent--long enough to burn more than 500,000 gal of supercold cryogenic liquid propellants. Developed in the 1970s by NASA's Marshall Space Flight Center in Huntsville, Alabama, the SSME is the world's most sophisticated reusable rocket engine.

Shown in Fig. 1, each engine has two powerful high-pressure turbopumps that supply up to 970 lb of liquid oxygen (oxidizer) per sec and up to 162 lb of liquid hydrogen (fuel) per sec to the engine's main combustion chamber. In this chamber, the propellants mix and burn at high pressure at temperatures exceeding 6000F (3516C) to produce thrust.

The 20-year-old turbopump machinery was constructed of a welded design (more than 300 welds of many separate but simplified forgings, castings and sheet metal parts) and required frequent maintenance between flights. This design proved to be timeconsuming and required extensive recycling, weld repair and reassembly, resulting in high total costs. These turbopumps required complete teardown and inspection after only one mission. Only then, upon a satisfactory report, were they reassembled for reuse.

The new Pratt & Whitney pumps, with a heavy reliance on castings, require fewer parts overall and 50% less rotating parts. The new turbopump's durability (10 missions) significantly increases the number of missions between major overhauls, which adds up to significant dollars, considering each of the three engines is valued at $45 million.

According to Len Worlund, director of advanced transportation at NASA's Marshall Space Center, the opportunity to address the welds and their inspection difficulties was what drove the engine redesign. "We knew that we could enhance the safety of the SSME if we could incorporate more robust, longer-life structures while eliminating the weld elements--particularly those that couldn't be inspected," he said.

Getting There

To fully optimize all areas of the engine, a multidisciplined group was formed with representatives from all aspects of design, analysis, manufacturing and procurement. "To reduce welds, we knew we needed more complex castings," said Mike Paytas, Pratt & Whitney project engineering manager. "In effect, we designed the pump around the castings. Because we were doing that, we were able to push the casting process technology envelope while also utilizing very complex configurations. Casting lead-time also was reduced because we worked in a truly concurrent fashion in the design process."

Because Howmet Corp., Greenwich, Connecticut, was the major casting supplier on the project and served on the integrated product team (IPT) from the "get-go," it could offer input on manufacturability and opportunities available through casting, said Terry Caulfield, Howmet senior technical representative. "It was unique because we became an integral part of their design effort and therefore had the opportunity to influence the ultimate part itself to make it more producible. We had a hand in designing fillets and other geometries to get the maximum properties needed to make the metalcasting process successful. The whole program benefited from the IPT structure working in the way it is intended."

The Pumps

In 1986, Pratt & Whitney accepted a challenge to develop alternate turbopumps for the SSME that would be capable of 55 missions between major overhauls, a factor 10-20 times better than the previous pump designs. In describing the thought that went into design, Paytas said, "A main philosophy was to reduce welds--they were always the weak point. Most cracks originated out of the welds, and by eliminating those weak spots, we were able to increase durability. We took components that were machined and/or cast in many different pieces and cast them in one piece."

The majority of the castings on the turbopumps were produced in innovative investment casting methods by various Howmet Corp. facilities. These investment castings provided complex designs that included volute flowpaths, hidden passages, massive flanges and internally cast airfoils and flow guides. The fuel pump also incorporated a highly cored aluminum sand casting produced at Fansteel/Wellman Dynamics, Creston, Iowa.

Liquid Oxygen Pump

The liquid oxygen pump, which was certified and flown in 1995, is currently certified for 10 flights--more than 5 times better than the previous liquid oxygen pump. According to Paytas, the pump is in the process of being extended to a 20-flight schedule, with the intent to further increase its cycle life. All the castings (see Fig. 2) on the liquid oxygen pump are produced by Howmet via investment casting in three innovative materials--Inconel 718, Mar-M-247 and PWA 1480 (see sidebar).

The Inconel 718 and Mar-M-247 alloys were advanced by Howmet's utilization of a proprietary, fine-grain casting process known as Microcast. In this modification to investment casting, the preheated shell-coated casting trees are positioned in vacuum furnaces. With a very narrow control of pouring temperatures (20F above liquidus), the process results in a homogenous fine-grain microstructure previously unavailable through casting. The fine-grain Inconel 718 alloy meets the SSME requirements for tensile strength, low cycle fatigue, cryogenic toughness and weldability. Meanwhile, the fine-grain Mar-M-247 process provides the turbine details needed with superior thermal strain at elevated temperatures and with low cycle fatigue capabilities.

The design also includes another high-tech casting technology-single-crystal castings (Fig. 3). Made from more advanced superalloy compositions of a higher melting point and greater strength, this single-crystal structure is produced by adding a single-crystal selector, or seed, to an investment cast mold. When the molten metal contacts a chill plate, the molten metal solidifies in a controlled manner that results in a casting with only one crystal. The superior attributes of these single-crystal components are primarily due to the absence of grain boundaries in the structure and the controlled crystal orientation. These advantages combine to provide extended operating lives of components--particularly in high temperature applications--thereby reducing the frequency of part replacement.

Following are the key castings on the liquid oxygen pump:

* forward and inlet housings (88 lb) and discharge housing (106 lb), produced in fine-grain Inconel 718. Welded together to form the main housing, these components serve as pressure vessels and flowpaths for the liquid oxygen propellant;

* turbine turnaround duct (26 lb), turbine inlet housing (35 lb) and first stage vane ring (5.2 lb), produced in fine-grain Mar-M-247;

* segmented vanes (0.5 lb), produced in fine-grain Mar-M-247, These, along with the turbine turnaround duct, inlet housing and first stage are hot system components that create structures and the flowpaths that guide hot gases through the turbines and into the turbine blades;

* 1st, 2nd and 3rd stage blades of single-crystal PWA 1480. At 1.0-1.5 in. long and weighing less than 1 oz, these parts turn the hot gas emerging from the preburner into mechanical energy, which in turn provides power to the pump's impellers.

The turbopump also includes a preburner volute casting and forward and rear shrouds, both cast in fine-grain Inconel 718.

The main difference on the new design, said Paytas, is that the components in the turbine inlet and turnaround duct are now one-piece vs. welded/fabricated parts. In addition, the vane castings featured many more airfoils that were cast to closer tolerances.

Crediting the fine-grain castings, Worlund shared that the welds were reduced from 300 on the existing design to only four main welds (seven in total) on the new design. Added Paytas: "We were able to achieve that because the casting configurations achieved were so complex."

Fuel Pump

In addition to the same materials and processes used on the liquid oxidizer pump (all the investment castings are produced by Howmet), the fuel pump also features an aluminum sand casting produced by Fansteel/Wellman Dynamics--the only sand-cast part on the SSME.

Key castings on the fuel turbopump include:

* pump inlet housing (200 lb), pump discharge housing (250 lb) and vane ring (37 lb), cast in fine-grain Inconel 718. It serves as pressure vessels that guide the flow for liquid hydrogen propellant;

* turbine inlet housing (44 lb), cast in fine-grain Mar-M-247. It is the flow guide and structure for the hot gas system;

* 1st and 2nd stage turbine blades, cast in single crystal PWA 1480. These lightweight parts power the pump impellers by turning the hot gases into mechanical energy;

* diffuser pump (80 lb) cast in aluminum A357 (Fig. 4). This part also serves as a pressure vessel component and flow guide for the engine.

In regard to the sole sand casting, Paytas said Pratt & Whitney took an already complex component and increased its complexity of the pencil cores to further refine the engine's performance. According to Fansteel/Wellman Project Engineer Alan Austin, there are 26 long and twisting cored paths in the front diffuser and 13 in the back that assist thrust balance and flow recirculation in the fuel pump. The two parts are produced in a coldbox sand mold with chemically bonded cores.

According to Worlund, all 469 welds that existed on the previous pump were eliminated with the new design. With this design, the mandated teardown after two flights is expected to increase to 10 when the project is certified.

Untapped Opportunities

With the Block II engine, the SSME. has an increased lifecycle (before teardown) of nearly 30%. "That translates into significant cost savings with reduced maintenance and reliability costs," said Worlund. "Inspections were usually on sheet metal cracks, and by getting rid of the sheet metal and replacing it with castings, we got rid of unwanted maintenance."

The two casting processes resulted in a turbopump design that has allowed:

* improved producibility;

* fewer welds;

* reduced manufacturing leadtime;

* mechanical properties comparable to wrought materials;

* reliable quality through rigorous process control and 100% inspectability;

* reduced cost;

* unsurpassed part repeatability.

"Castings still are not perfect," said Jeff Bland, NASA residence officer at Pratt & Whitney. "They can have little imperfections that can require some clean-up, and there are certain applications where they can be difficult to inspect. But their big plus is the geometries that they can achieve. Some of the volutes on this engine are extremely detailed. Castings can allow some very intricate designs."

Caulfield concluded that the power of castings is yet to be tapped, largely because of the unfamiliarity of castings or the trap of yesterday's paradigms on the metalforming method. "Some engineers still mistakenly think of castings as a debit, and as a result, castings are sometimes looked at as the last options rather than one of the first," he said. "They really must keep an open mind to the possibilities. Some of the applications being done today, such as those like the SSME, are absolutely amazing."

Unique Cast Materials Used on the SSME Turbopumps

Inconel 718--a nickel-based super-alloy with nominal composition of 62% nickel (Ni), 19% chromium (Cr), 9% iron (Fe), 5% columbium (Cb), 3% molybdenum (Mo), 0.9% titanium (Ti), 0.6% aluminum (Al) and 0.05% carbon (C).

MAR-M-247--a heat resistant Ni-based alloy with a nominal composition of 59% Ni, 10% cobalt (Co), 10% wolfram (W), 8.4% Cr, 5.5% Al, 3% tantalum (Ta), 1.5% hafnium (Hf), 1% Ti, 0.7% Mo, 0.15% C.

PWA 1480--A proprietary single-crystal alloy with a composition of 63% Ni, 12% Ta, 10% Cr, 5% Co. 5% Al, 4% W and 1.5% Ti.

Fastrac--NASA's Low-Cost Launch Vehicle Engine

In a separate project, NASA Engineers at the Marshall Space Flight Center are designing what may be one of the world's simplest turbopump rocket engines. The new Fastrac engine is, true to its name, on a fast track to propelling the next generation of launch vehicles, which will put research payloads into orbit at a much reduced cost than currently possible.

The new Fastrac engine is 7 x 4 ft, and weighs nearly 2000 lb--less than half the size and about 33% the weight of the SSME. With an individual engine cost of about $1 million, the Fastrac is about 14% more cost-effective than similar engines. The first vehicle to be powered by the new engine is the X-34, a technology test bed vehicle launched from the underside of a L1011 aircraft. This vehicle will be used to demonstrate key vehicle and operational technologies applicable to future lowcost reusable launch vehicles.

Technology and development and design for the Fastrac engine began in early 1995. In a salient departure from traditional engine design, NASA and its partners utilized commercial, off-the-shelf technologies and common manufacturing methods-including cast metal components-to develop the Fastrac engine. To cut the turbopump cost by nearly the 33% requirement, significant small business nontraditional contractors were used to produce the lower cost hardware.

As an example of the new process, Barber-Nichols, Inc., Arvada, Colorado, a turbopump supplier for the auto and chemical industries, was selected to partner with Marshall engineers to design and manufacture the turbopump. The company contributed a turbopump design that can be reliably and cost-effectively produced using commercial manufacturing techniques such as metalcasting.

With low costs as a prime requirement of their mission, Barber-Nichols and NASA focused their turbopump design on the use of three Inconel 718 metal castings (a33-lb liquid oxygen pump, 65-lb fuel pump and 40-lb turbine inlet), primarily to consolidate components and reduce weldments. Both investment casting and sand casting (via nobake molds) are currently being evaluated for the turbopump castings. In fact, the sand-cast technique for the superalloy (which is not typically sand cast nor air-melted) was developed further as a result of this process, and is showing an overall cost reduction of 71%.

Jim Dillard, a design engineer with Barber-Nichols, has been designing turbopump machinery for nearly 30 years. Casting has become his primary method of choice when designing turbo machinery, but this wasn't always the case. "In the years past, we found that it could take a long time to get parts from our casting suppliers," he said. "Today, things have changed--foundries can supply parts faster than the other alternative processes. With the implementation of rapid prototyping and CNC machining in pattern shops, we can have production parts in a few days and not weeks or months."
COPYRIGHT 1999 American Foundry Society, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:ROSMAIT, R.L.
Publication:Modern Casting
Geographic Code:1USA
Date:Dec 1, 1999
Words:2559
Previous Article:CERP: Helping Foundries Identify Routes to Reduced Air Emissions.
Next Article:Speakers Update Foundrymen on 'Hot' Regulatory Issues.
Topics:


Related Articles
ME Global helps launch shuttles to space: the track shoes on NASA's crawler transporters, which carry the space shuttles to the launch pad, required...

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |