Motion Technology Powers Planetary Exploration
Motion designers leveraged innovative technology and the careful design and testing to power Curiosity, the latest Mars rover.
By: Kristin Lewotsky, Contributing Editor
Harsh environments challenge motion control in a wide range of applications. Of course, there are harsh environments and then there are harsh environments. The conditions encountered on a factory floor, even one entailing wash-down operations, are very different from those encountered in a down-hole drilling for oil and natural gas application or even on a battlefield in high summer in Iraq. When you take things out into space, these become even more extreme.
As petroleum reserves dwindle, companies are drilling deeper and deeper to find oil. Servo motors can help in down-hole measure-well-drilling (MWD) tools. The motors operate valves that relay signals to the surface to identify drilling location and depth. The instruments are held in a tubular enclosure attached to the drilling tool itself. By any standards, conditions are hostile, subjecting the motion components to shock and vibration as well as temperatures that are increasing to around 200° C as the tools drill deeper. "When you’re talking about harsh environments, that‘s about as extreme as it gets," says Mark Dwyer, Sales Engineer at maxon precision motors, inc. (Fall River, Massachusetts). The application does not require high speeds, but the motors need to supply a significant amount of torque.
The temperatures require the use of exotic magnets such as samarium cobalt, which can withstand temperatures of up to 300°C. Connecting motor and gearbox by laser welding can render the device less vulnerable to shock-induced failure. Interestingly, the instruments operate filled with a low-viscosity oil that actually helps the motors withstand the high pressures produced at such depths.
The tools integrate an external control board housed near the motor. Currently, the temperatures for most down-hole applications remain around 175°C, which makes it easy to find control and drive electronics. The difficulty, Dwyer says, will arise as companies drill deeper and down-hole temperatures begin to soar to 200°C or higher. "It's challenging because there aren't many electronic components out there that can handle 200°C, and high vibration, and operate flooded in oil," he says. For now, MWD vendors have been developing their own technology, which satisfies current conditions.
Down-hole drilling may pose some of the toughest terrestrial conditions, but some of the harshest environments of all are found in space. Temperatures range from negative hundreds of degrees centigrade to positive hundreds of degrees. Components are punished by a bombardment of ionizing radiation. The launch process subjects them to high degrees of shock and vibration. Despite all of these factors, performance requirements remain high.
One means to creating a vibration-tolerant device is to coat the board so that components cannot move and can withstand high or extended vibration. This type of coating brings big benefits, but it does also limit options. "The upside of that is that it protects the electronics, it covers the whole board," says John Hayes, Senior Applications Engineer at Galil Motion Control. “Once you do that, though, you cannot really work on the board. So the downside is that once you do it you are not going to be able to fix it or make any changes." Fastening components with an adhesive like Loctite can also improve robustness in the face of vibration.
Although motion components designed for orbital or deep space flight may face challenging conditions, there are perhaps none more difficult than those presented by the newest Mars rover, Curiosity, currently heading to Mars on the recently-launched Mars Science Laboratory. Approximately 2.7 m x 3 m x 2.1 m tall, the rover weighs roughly 900 kg (2000 lbs). Specifications call for it to travel at speeds up to 90 m/hr on the surface and navigate over objects as large as 75 cm (29 in) high. Fully extended, the articulating arm stretches to 2.2 m.
In all, the rover incorporates 32 axes of motion. They include a drive and steering motor on each wheel; plus motors to articulate the arm; the sample, acquisition, and processing system; and the remote-sensing mast. Additional motors support the deployment process, opening covers and so on. The only diagnostic data available once the unit reaches Mars will be current traces for each of the actuators, defined by the team as motor, gearbox, and encoder.
Although Spirit and Opportunity both had multiyear lifetimes with brushed motors, for Curiosity, the team went with brushless designs. The choice was driven primarily by the high shock and vibration environment that the rover needs to endure during launch, flight, and deployment. The mission calls for Curiosity to make much of the trip from orbit to the Martian surface by parachute, then to be lowered on tethers for the final segment. “It's a high shock environment," says Kobie Boykins, Group Supervisor For Mobility and Mechanisms at the Jet Propulsion Laboratory and Actuator Cognizant Engineer for Curiosity. "During descent, there are a lot of things going on I call reverse origami. We’re releasing things, we’re popping parachutes, we’re deploying mobility--the rover starts to unfold. All those releases or shock events tend to be near motors and gearboxes." Brushes tend to be susceptible to shock. In addition, the Martian atmosphere is high in carbon dioxide, which also contributes to brush wear. That, combined with incompletely understood plasma dynamics taking place between rotor and brushes drove the team to select a brushless design.
In order to use the brushless motors, the team needed encoders. To conserve power, they developed a magnetoresistive design capable of working at temperatures down to -128°C. The next focus was to try to develop gearboxes and bearings designed for these types of temperatures. Although lubricant is essential, especially in arid environment, wet lubricant posed the risk of freezing. The team began development on a low-temperature gearbox.
A project like the Mars Science Laboratory absolutely requires an extremely robust system. After all, there is no "do over" for a robot deployed tens of millions of miles away. That means designing for reliability and testing, testing, testing. The development of the low-temperature gearbox, a challenging task, proved to be a somewhat star-crossed effort. Initial testing resulted in failure. The cause ultimately proved to be a commercial, off-the-shelf bearing not designed for low-temperature operation. It was part of the test setup, not part of the gearbox itself. Unfortunately, its malfunction triggered catastrophic failure. "The bearing had a retaining ring that was not really approved for doing temperature testing," Boykins says. "We started low-temperature testing and that retainer ring got ingested into the gearbox and shot through it. That stopped that test," he adds, unnecessarily.
Of course, in the immediate aftermath, they could make no assumptions. Instead, they spent several months studying the incident to determine the proximate cause of the failure. With the replacement bearing, they launched back into testing. This time, too, testing terminated early because of failure, this time as a result of ground support equipment (GSE), not the gearbox. Once the team had diagnosed the problem and begun again, things appeared to be going promisingly, until the telemetry data began to exhibit strange artifacts. Upon taking apart the gearbox, the group found problems once again, despite the fact that they had corrected issues with the gearbox and the GSE. At this point, given the project timeline and the challenges of the development effort, the reviewing team recommended temporarily sidelining the cold gearbox idea in favor of a conventional technology.
The motors had to run at -135°C for hundreds of millions of revolutions. As testing proceeded, the bearings on different axes began displaying intermittent instabilities. In tests under no-load conditions, the team noticed variations in the current telemetry on the order of 100%.
Bearing instabilities can stem from a number of different causes. The bearings may skate or slide instead of rolling. The cage or separator can lock up with the balls inside the bearing. The retainer and the inner or outer rings can rub. Instabilities can even stem from rotor dynamics because of the speeds at which the motor is turning. “You get to a fundamental frequency that turns your device into the Tacoma Narrows bridge, if you will," Boykins says. Of course, if finding the source of bearing instability is difficult, correcting it is even more challenging. "In the industry it’s sort of a dark art. People take out a ball inside the bearing, they change something else. There really is no one answer. You’re in a world of experimentation to get to the point where you get a quiet system.”
The team struggled to understand the issue and identify the cause. In the case of the low-temperature gearbox, the team was able to make the decision to return to a conventional technology. In the case of the bearing instabilities, they were not so fortunate, given that the effect manifested for both new technology and standard versions that had been in use on various projects for decades. Unfortunately, there was no easy answer, outside of ensuring that the overall design was robust enough and had sufficient performance overhead to tolerate the occasional instabilities.
That wasn't the end of the development battle. After one round of tests, they received current telemetry that indicated a sinusoidal load on the gearbox, which contradicted to the constant-torque that the team was applying to the system. When they investigated, they discovered that the input bearing of the gearbox had fractured in a failure that redefines the term catastrophic. The heart of the problem was preloading. In an attempt to save size and weight in the spacecraft and simplify the design, the plan calls for the rover’s wheels to absorb the shock of contacting the planetary surface. It was a worthy idea in theory but in practice it caused the external landing loads to pass through the output bearings of the actuator.
In order to ensure survival of the components, the team had preloaded the output bearings on the gearbox. The problem was that the anticipated moment loads from landing were too large for the bearing configuration; hence, catastrophic failure. “These are angular bearings," Boykins says. "The preload was not such that they could actually take the moment. The load path was different than what we had originally believed. That’s how the load went through and broke the bearing, because now the actual stages of the gearing were taking the load, as opposed to the housing carrying the load to the outside of the housing and down to the feet." After exhaustive analysis, with a looming deadline, the team redesigned the preloading process to direct loads over a different path.
And, they started their lab tests again.
That wasn't entirely the end of the challenges. The rover's articulated arm subsequently caused problems with the gearbox as a result of the weight of the components and the length of the moment arm. Again, the unit failed in testing. Again, the project assembled a review team that analyzed the situation. This time, after investigating a range of options, testing to failure, and analyzing the results, the team determined that there was no way to accomplish the required goals within the existing set of constraints. Something had to change, whether that involved performance, cost, or size.
"You have a large moment on the length of the arm, so that makes a large moment coming back into the gearbox," says Boykins. "Now, we have to come back and say, ‘Okay, we need to lower our stress in these parts so that we will not be at the fatigue limits and we can actually last for the lifetime." This time, the equipment successfully passed the lifetime tests, and the team was ready to move forward.
Mars Science Laboratory launched on November 26, successfully entering its trajectory toward Mars. As of this writing, all communications tested out properly. Plans initially called for a controlled burn on December 10 to correct course, but it was postponed after engineers determined that the spacecraft is on trajectory. It is scheduled to land at the Gale Crater on August 6 two began its 686-day mission
For Boykins, who also worked on Spirit and Opportunity, watching the rovers land is another highlight in an extraordinary project. "It’s the most amazing feeling," he says. "You’re just overwhelmed with joy and with respect for all of the people that you’ve gotten a chance to work with. It’s humbling to be part of building something that is going to land on another planet. It is just the coolest feeling ever.”
Thanks go to Jeff Randall of maxon precision motors for useful conversations.