# FRC Drive Train Design and Implementation

AI and Robotics

Nov 2, 2013 (4 years and 8 months ago)

121 views

2008
FIRST

Robotics Conference

FRC Drive Train Design and
Implementation

Presented by:

Fred Sayre, Team 488

Who are we?

What is a drive train?

Reexamine their purpose

What won’t I learn from this presentation?

No use reinventing the wheel, so to speak

Why does that robot have 14 wheels?

Important considerations of drive design

Tips and Good Practices

All in 40 minutes or less. We hope.

2008
FIRST

Robotics Conference

2008
FIRST

Robotics Conference

Who Are We?

2008 is 10
th

season with FIRST

Lead Design Mentor for Team XBot

Fred

2008 is 6
th

season with FIRST

What is a drive train?

Components that work together to move robot
from A to B.

Focal point of a lot of “scouting discussion” at
competitions, for better or
for worse
.

It has to be the most reliable part of your robot!

That means it probably should be the least

unless you’re
awesome.

2008
FIRST

Robotics Conference

This presentation is not…

a math lesson.

Ken Patton’s presentation will rock your world.

a tutorial.

you can do, so there is no single solution that is best
for all teams

unbiased.

We call it like we see it. Your mileage may vary.

2008
FIRST

Robotics Conference

Why does that robot have 14 wheels?

Different field surfaces

Inclines and steps

Pushing or pulling objects

Time
-

Omnidirectional motion is useless in a drag race

but great in a minefield.

2008
FIRST

Robotics Conference

Important Concepts

Traction

Double
-
edged sword

Power

More is better?

Power Transmission

This is what makes the wheels on the

bus go ‘round and ‘round.

Common Designs

2008
FIRST

Robotics Conference

Traction

Friction with a better connotation.

Makes the robot move

Keeps the robot in place

Prevents the robot from turning when you intend
it to

Too much traction is a frequent problem for 4WD
systems

Omniwheels mitigate the problem, but sacrifice some
traction

2008
FIRST

Robotics Conference

Power

Motors give us the power we need to make things
move.

Adding power to a drive train increases the rate at
which we can move a given load
or

increases the
load we can move at a given rate

Drive trains are typically not “power
-
limited”

Coefficient of friction limits maximum force of
friction because of robot weight limit.

Shaving off .1 sec. on your ¼
-
mile time is meaningless
on a 50 ft. field.

2008
FIRST

Robotics Conference

More Power

Cooler motors

Decreased current draw; lower chance of tripping
breakers

Redundancy

Lower center of gravity

Drawbacks

Heavier

Useful motors unavailable for other mechanisms

2008
FIRST

Robotics Conference

Power Transmission

Method by which power is turned into traction.

Most important consideration in drive design

Fortunately, there’s a lot of knowledge about
what works well

Roller Chain and Sprockets

Timing Belt

Gearing

Spur

Worm

Friction Belt

Power Transmission: Chain

#25 (1/4”) and #35 (3/8”) most commonly used in
FRC applications

#35 is more forgiving of misalignment; heavier

95
-
98% efficient

Proper tension is a necessity

1:5 reduction is about the largest single
-
stage
ratio you can expect

Power Transmission: Timing Belt

A variety of pitches available

Frequently used simultaneously as a traction
device

Treaded robots are susceptible to failure by side
-

Comparatively expensive

Sold in custom and stock length

breaks in the
belt cannot usually be repaired

Power Transmission: Gearing

Gearing is used most frequently “high up” in the
drivetrain

COTS gearboxes available widely and cheaply

Driving wheels directly with gearing probably
requires machining resources

Spur Gears

Most common gearing we see in FRC; Toughboxes,
NBD, Shifters, Planetary Gearsets

95
-
98% efficient per stage

Again, expect useful single
-
1:5 or less

Power Transmission: Gearing

Worm Gears

Useful for very high, single
-
stage reductions (1:100)

Difficult to backdrive

Efficiency varies based upon design

anywhere from
40%

Design
must

Power Transmission: Friction Belt

Great for low
-
friction applications or as a clutch

Apparently easier to work with, but requires high
tension to operate properly

Usually not useful for drive train applications

Common Drive Train Styles

Skid Systems

2WD, 4WD, 6WD, 6WD+

Holonomic Systems

Swerve/Crab

Mecanum

2008
FIRST

Robotics Conference

Two Wheel Skid | Four Wheel Skid

The Good

Cheap;
Kitbot

is 2WD

Very simple to build

Easily spins out

Difficulty with inclines

Loses traction when
drive wheels leave floor

The Good

More easily controlled

Pretty simple to build

Better traction

Turning in place is
more difficult

Compromise between
stability and
maneuverability

2008
FIRST

Robotics Conference

6 Wheel Skid

Typically, one wheel is offset from the others to
minimize resistance to turning

Rocking creates two 4WD systems, effectively

Typical offset is 1/8”

¼”

Rock isn’t too bad at edges of robot footprint, but can
be significant at the end of long arms and appendages

One or two sets of omniwheels can be substituted
for offset wheels.

2008
FIRST

Robotics Conference

In the real world, we’d add more wheels to
distribute a load over a greater area.

Not a historically useful concept in most FRC games,
Maize Craze possibly being an exception

Simply speaking, traction is not dependent upon
surface area

Deformation plays a role in reality

Diminshing returns

Mechanically complex and expensive for marginal
return

2008
FIRST

Robotics Conference

Holonomic Drive Systems

Allow a robot to translate in two dimensions and
rotate simultaneously

Two major mechanical systems

Swerve/Crab

Mecanum/Omni

2008
FIRST

Robotics Conference

Holonomic Drive Systems:
Swerve/Crab

Naming isn’t standardized. I use them
interchangeably.

Most FRC drives of this type are not truly
holonomic

That requires wheels that are driven and steered
independently

Holonomic Drive Systems:
Mecanum/Omni

Uses concepts of vector addition to allow for true
omnidirectional motion

No complicated steering mechanisms

Requires four independently powered wheels

COTS parts this system accessible to many teams

Tips and Good Practices

KISS

Keep it Simple, Stupid

We’re trying to get RRRR into the lexicon

Reliability

Reparability

Relevance…ability

Reasonability

Tips and Good Practices: Reliability!

Most important consideration, bar none.

Three most important parts of a robot are, famously,
“drive train, drive train and drive train.”

Good practices:

Support shafts in two places. No more, no less.

Avoid press fits and friction belting

Alignment, alignment, alignment!

Reduce or remove friction almost everywhere you can

Tips and Good Practices:
Reparability!

You will probably fail at achieving 100% reliability

Good practices:

Design failure points into drive train and know where
they are

Accessibility is paramount. You can’t fix what you
can’t touch

Bring spare parts; especially for unique items such as
gears, sprockets, transmissions, mounting hardware,
etc.

Aim for maintenance and repair times of <10 min.

Tips and Good Practices:
Relevance…ability…!

Only at this stage should you consider advanced
thingamajigs and dowhatsits that are tailored to
the challenge at hand

Stairs, ramps, slippery surfaces, tugs
-
of
-
war

Before seasons start, there’s a lot of bragging
about 12 motor drives with 18 wheels; after the
season is over, not as much

Tips and Good Practices:
Reasonability!

Now that you’ve devised a fantastic system of
linkages and cams to climb over that wall on the
field, consider if it’d just be easier, cheaper,
faster and lighter to drive around it.

FRC teams

especially rookies

grossly
overestimate their abilities and, particularly, the
time it takes to accomplish game tasks.

Resources

ChiefDelphi

Internet forum watched by the best of the best

A lot of static, but patience yields great results

http://www.chiefdelphi.com

FIRST Mechanical Design Calculator by John V
-
Neun

http://www.chiefdelphi.com/media/papers/1469