Welding Capabilities and Technologies at Marshall Space Flight CenterJeffery Ding - NASA's Marshall Space Flight Center
Mr. Ding has been deeply involved with developing improved welding techniques for the 10000 or more welds in a main shuttle engine alone, as well as those for various tanks and other components. At Marshall SFC there are three main divisions:
Systems - Programs/Projects
Space Transportation - Programs/Projects
Shuttle Propulsion
The Materials and Processes Laboratories (Mr. Dings's home) has the purpose of advancing the state-of-the-art of metals processes.. etc. Recent capabilities that have been developed here include (primarily in the area of "solid-state" welding):
Friction Pin Tool Technology
Retractable Pin Tool Technology
Plug Repair Technology
High Speed Friction Surface Welding
Self Reacting Pin Tool Technology
and Thermal Stir Processes (independent control of heat, stir, and forge)
As background, during the periods of 1975-82, most welding on NASA projects was either TIG (tungsten inert gas), EB (Electron Beam), and limited Inertia Welding, with TIG being the main process.
Beginning in 1982, Variable Polarity Plasma Arc Welding was developed and was a vast improvement over the TIG process. For aluminum welding of external tanks, an 8 pass process (TIG) was replaced with a 2 pass (penetration / cover) process (using VPPA) with a great reduction in defects.
Since the mid 1990's the move has been toward "Solid-State" welding which offers several advantages. This is a non-melting process which eliminates dendrites in the microstructure. The objective is to bring the metal to a plastic state, where it flows together, dynamically recrystallizes, and returns to near normal (non-heated) properties. Since no filler material is used, there is some loss of cross-sectional area and strength associated with this process, but despite this -- the solid-state welds are generally much stronger than when a traditional dendritic and heat effected zone are created by fusion welding.
The first form of this process was developed at the Cambridge Weld Institute (UK), and Marshall Spaceflight Center has been perfecting the process ever since.
A Friction Stir Weld is accomplished by using a tool with two components. A pin (with a diameter somewhere in the range of weld metal thickness, is designed so that when spinning it agitates (by being fluted or threaded). This "agitator" progresses down the line where the two weld surfaces are held together, and causes the mixing and dynamic recrystallization of the plastic-state metal in its path. To get the metal heated to the point where this is possible, the pin is surrounded by a collar (1" to a few inches in diameter) which is driven against the top flat surface of the weldment by considerable force (hundreds to thousands of pounds), causing high friction and heat as it rotates along with its pin. This causes major preheating of the weld line, and allows the pin to do its job. The pin itself usually extends to within a few thousandths of an inch of going through the thickness of the welded plate.
For low melting weldments (ie. aluminum), pin a collar can be simply a single fixture made of tool steel. For higher melting weldments, higher melting superalloys are required for tools, and for welding "high temperature alloy" metals, the collar and pin may be separate materials, including refractory metal alloys like Mo-Re.
Considerable development has determined precise rotation speeds, collar pressure (which must be backed up by a very rigid anvil on the back side of the weld), and feed speeds along the weld line to get very reliable welds. Due to the high collar pressure, welds are usually a few percentage points thinner than the plates being welded. Typically tool rotation is in the area of 200 - 400 rpm with applied force of 600 - 14000 lb (for 1" thick material), and travel along the weld line is 4 - 5 inches per minute.
One problem with this method is that at the end of the weld a hole is left by the pin of the pin tool. In 1996 this was eliminated by developing a "retractable pin tool", which (being a separate piece within the collar) could be retracted slowly at the finish of the weld, allowing the hole to fill in with plastic metal.
By 1999 the process was accomplishing circumferential welds (on large diameter tanks).
A more recent development has been the "self-reacting friction stir weld". In this process, a rotating collar is applied from both sides of the plates to be welded (eliminating the anvil). The pin is still a stirring member which is coaxial to both collars and passes down the weld line. A main advantage of this method, is that a "lack of penetration" defect (where the pin did not stay close enough to the bottom of the weldment) is eliminated. With this process the pin agitates completely through the weld thickness. Additionally there is no hole at the end of the weld to worry about. Full scale self-reacting friction stir welding has been demonstrated on a 27.5 foot diameter tank surface.
Beyond these developments, the processes have been developed and tooled to do complex curvature welds but the weld tools for these are very expensive.
Currently a project is underway to meet the need for "in-space" maintenance (including weld repair), because current moon and Mars projects will require "Fabrication on the Fly". It is hoped this can be accomplished with Friction Stir Welding (much safer that dealing with molten metal spatter in space). The plan is to get the heat required with much higher rotation speeds (1000 - 30000 rpm) while reducing the forces on the tool to what can be applied by an astronaut. Packer Technologies Inc. is working on a prototype for such a "space walk" weld tool.
Finally, work is underway to developing a more flexible variation of solid state welding where heating is independent (say ultrasonic), stirring is still a pin, and forging is an "immediately after stir" separate operation. By being able to control these three parameters separately, much more flexibility can be achieved for the process.
The Technology Transfer arm of NASA is always ready to offer any of its processes for co-development. The officer to contact concerning these processes is Sammy Nabors or Marshall Spaceflight Center.