2008 Drive-train Development

From DEW Robotics
Jump to: navigation, search

Due to several other challenges in the 2008 build season, DEWBOT IV's drive-train largely became an afterthought. The system was a severely underwhelming chain-driven rear-wheel drive that did not steer straight, broke regularly, and tended to induce wheelies until the center of gravity was changed. However, the team chose to grow from this experience by spending the subsequent summer researching, prototyping, and analyzing drivetrains. This study was summarized in a PowerPoint presentation by Clem McKown finished in August of 2008.
The results of this endeavor were the Provisional DEWBOT V Chassis, which was designed before the 2009 season (before the low-friction regolith was introduced) and the actual DEWBOT V Drive Train, which was the main reason the team won its first regional award, the Rockwell Automation Innovation in Control Award.

Designs Considered or Tested

Mecanum

Twitch

6wd

Specific Lessons Learned

See also Weaknesses of DEWBOT IV

Undriven wheels

DEWBOT IV Problem: Rear-wheel drive (with two undriven wheels) significantly reduced traction and caused other problems.
Lesson: Never have undriven wheels! The possibly simplification on the drive train is not worth it, especially for wheels that support significant weight.
DEWBOT V Correction: DEWBOT V had six-wheel and three-wheel configurations with all wheels were driven in both, despite requiring an extra motor for the perpendicular 7th Wheel. This design helped the team win the Rockwell Automation Innovation in Control Award at the Chesapeake Regional.

Uneven Drive Base

DEWBOT IV Problem: An uneven drive base meant one wheel was often unloaded (as three points, not four wheels, define a plane). This greatly impeded steering.
Lesson: Keep the drive base level! (Also, drive of the wheels to negate the effects of any loss of contact.)
DEWBOT V Correction: Despite it being the team's foray into welding, DEWBOT V's frame was very balanced and did not bend significantly. (Also, it had either three of six, all driven wheels.)

Unbalanced Center of Gravity (Steering and Instability)

DEWBOT IV Problem: Originally, the rearward center of gravity and rear-wheel drive induced numerous wheelies. However, bringing the center of gravity further weighted the undriven front wheels and exacerbated the reduced traction and steering problems.
Lesson: Pay attention to the center of gravity! Keep it low and centered and consider how it will impact driving and other functions.
DEWBOT V Correction: DEWBOT V's center of gravity was kept very low (9"?) and almost exactly centered (3" back?). The robot never experienced wheelies and was able to push other robots even on the slippery regolith. It was also incredibly stable, even in its tilted three-wheel configuration.

Difficult Drive-train Maintenance

DEWBOT IV Problem: Due to the extreme tension and poor design of its drive train, DEWBOT IV broke chains very regularly--sometimes every match. Though the problem was later mitigated, the stigma against chains remained.
Lesson: Keep the drive train strong and secure and easy to maintain and repair. The summer also relieved the team of their aversion to chains.
DEWBOT V Correction: DEWBOT V's drive train is a work of art. The system is modular and the sides are identical. The middle wheels are direct-drive from custom gearboxes, and the outer ones are slave driven via chains. The sides are symmetrical No chain has ever broken, and only one lost tension once.

Difficult Wheel Tread Maintenance

DEWBOT IV Problem: With steering that required wheel slippage (especially unfortunate in a driving-in-circles game), DEWBOT IV's wheel treads wore out very quickly. They were riveted on, making them very difficult to replace and leading to complications during competition.
Lesson: Minimize wheel slippage and use easily replaceable treads/wheels.
DEWBOT V Correction: DEWBOT V's 6-wheel drive turns more easily (and slips less). It also incorporates a novel 7th Wheel mounted perpendicularly to assist in agile pivoting. Before rover wheels were introduced for the Lunacy game, the team found AndyMark Plaction Wheels (R) which easily come apart for tread replacement. These wheels were incorporated into the DEWBOT V Provisional Chassis and will likely be used in later games (in accordance with FIRST game rules).

Difficult Bumper and Battery Installation and Removal

DEWBOT IV Problem: No thought at all was given bumper removal and installation on DEWBOT IV. This could take upwards of forty-five minutes and caused significant problems at competitions. To improve the center of gravity, the battery nested awkwardly and haphazardly within the robot. It made for difficult and sometimes painful installation and removable and occasionally fell out, costing entire competition matches.
Lesson: Plan for bumpers, batteries and other often-removed devices! Keep their mounting safe, secure, easy, and quick.
DEWBOT V Correction: DEWBOT V's bumpers can be installed and removed in under a minute without any tools. The battery is mounted securely and diagonally on the frame. The current battery switch record is 12 seconds.