FENAL REPORT

June 1970

DESIGN AND FABRICATION OF WHEELS FOR A

LUNAR SURFACE VEHICLE

VOLUME 1
Detailed Technical Report

Prepared for
GEORGE C. MARSHALL SPACE FLIGHT CENTER
HUNTSVILLE, ALABAMA 35812
CONTRACT NO. NAS 8-25193

GRUMMANAERIolSPAGE G(ElRP(ElRAU0aN
BETHPAGE, NEW YORK 11714
 TABLE OF CONTEmS

VOLUME I
1 INTRODUCTION
2 DESIGN CRITERIA
2.0 General
2.1 Dynamics Criteria
2.1.1 Response to Random Terrain
T, 2.1.2 Loads Due to Obstacle Encounter
I 2.1.3 Selection of Wheel and Suspension Characteristics
2.1.4 References for Section 2.1
2.2 Loads and Structural Design Criteria
2.2.1 Stowed Condition Loads
2.2.2 Deployment Loads
2.2.3 Lunar Operation Loads
2.3 Environmental Criteria
2.3.1 Ground Environments
2.3.2 Lunar Environment
2.4 Mobility Considerations
2.4.1 Material Selection
2.4.2 Wheel Preliminary Design
2.4.3 Mobility Performance
2.4.4 Rolling Resistance
2.4.5 Slope Climbing
2.4.6 References f o r Section 2.4
2.5 Summary of Dynamic Design Criteria
3 DESIGN DEVELOPMENT AND SEIZCTION
3.0 General
3.1 Structural Analysis
3.2 Prior Grwnman Testing
3.2.1 Strain Gage Instrumented Test Wheel
3.2.2 Test Set Up
3.2.3 Test Results
3.2.4 Evaluation of Results
3.2.5 Conclusions
3.3 Wheel Design "A"
3.3.1 Fabrication Process
393.2 Test Plan
3.3.3 Test Results
3 04 Design Iteration
 TABLE OF CONTENTS (Continued)

3.4.1 Test Results

3.5 Final Design Selection
3.5.1 Final Design Characteristics
3.5.1.1 Spring Rate
3.5.1.2 Stress Distribution
4 FABRICATION PROCESS
4.0 Fabrication Requirements
4.1 Test Wheels
4.2 Deliverable Wheels
4.3 Cleat Fabrication
4.4 Wheel Weight

VOLUME I1

PROPOSED TEST PLAN

1.0 Purpose
2.0 Description of Test Article
3.0 Test Conditions
3.1 Static Tests
3.2 Dynamic Tests
3.2.1 Traffic Mobility
3.2.2 Ambient Endurance Tests
3.2.3 Thermal/Vacuum Tests
4.0 Test Fixtures
4.1 Wheel Suspension
4.2 Road Bed
4.3 Instmentation
4.3.1 Static Testing (Wheel)
4.3.2 Mobility Testing (Wheel)
4.3.3 Dynamic Endurance Testing
4.3.4 Static Testing (Drive Unit)
4.3.5 Dynamic Testing (Drive Unit)
5.0 Test Procedures
5.1 Establishment of Initial Test Con1 it .ons
5.1.1 Static Measurements
5.1.2 Load Increments
5.1.3 Obstacles
5.1.4 Side Loads
5.1.5 Inspection
 TABLE OF CONTENTS (Continued)

5.2 Mobility Testing

5.2.1 Suspension
5.2.2 Soil
5.2.3 Measurements
5.2.4 Inspection
5.3 Endurance Testing (Partial Vacuum)
5.3-1 Setup
5.3.2 Test Segments
5.393 Inspection
5 04 Endurance Testing (Full Thermal Vacuum0

i
 1 INTRDDUCTION

Long before man had set foot on the lunar surface the requirements for a
powered surface transportation vehicle had been established. Prior to the
success of Apollo 11, however, little quantitative data were available to
accurately define the mobility effecting characteristics of the lunar
surface. Ten years ago when the first lunar vehicles began to evolve, they
were based on the assumption that only the most lightly loaded vehicles
could operate on the soft homogenous surface. Later, data from the Ranger,
Orbiter, Surveyor and now the Apollo program have incrementally increased
our understanding of the properties of the lunar surface. As the data
1.1 bank of information grew, proposed vehicle concepts were altered to meet
the revised missions and mobility criteria. Grwamans Molab design
(shown in prototype form in Figure 1-1) with large metalastic wheels
typically satisfied the early mobility criteria.
As our knowledge of the lunar surface grew in detail, the concept of a
homogenous surface gave way to a debris strewn surface with large quantities
, '? of widely dispersed rocks in the 2-4 inch size. Initial vehicle mobility
system concepts were reevaluated and discarded because of the possibility
of debris entrapment within the wheel geometry, leading to ultimate wheel
failure.
At Grurmnan a single element elastic conical wheel (Figure 1-2) was evolved

I
"ra
I
through an extensive company Punded design and test program addressed to
resolving the lunar mobility problem. The results of full scale wheel and
model tests in addition to Gmumnans f u l l scale LRV simulator attest to the
success of this concept.
The objective of this study was to refine the existing conical wheel design
in order to provide an optimum configuration compatible with the NASA defined
Dual Mode Lunar Roving Vehicle (DLRV) mission requirements. The unique
wheel form does not allow the normal analytic analyses techniques therefore

i
i
t
?
c.
 an empirical test program was used to evolve an optimum design. Because of
the interrelated nature of the previous testing and the DLRV concepts, prior
data which was felt to be pertinent t o the final design selection shown
herein has been included in this report. Three wheels of the final design
have been provided to NASA for testing. A proposed lunar qualification
test plan is provided for the entire wheel drive assembly.
f

I-

J
 2 DESIGN CRITERIA
2.0 GENERAL
The design criteria used to develop the final wheel configuration was based on
the DLRV mobility requirements which defined static and dynamic loadings, and
the non mobility requirements of earth launch and lunar operations. These
criteria were documented in detail in GIZUIIII~E~I?'SDLRV final report, excerpts of
which are incorporated below.
2.1 DYNAMICS CRITERIA
Wheel design criteria are developed by assessing the interaction of the wheel
suspension system, the vehicle itself and a lunar soil model. The following

-7 discussion shows how the surface model, loads and vehicle motions are related,
and the rational for selection of the wheel-suspension configuration.
Vehicle/wheel dynamics studies were performed to optimize ride qualities, maximize
vehicle controllability (related to the amount of time the wheels are off the
ground), and determine damping power dissipation, turning stability boundaries,
and dynamic loads during lunar operation.
The dynamics effort consisted of three main areas: analog computer studies of
the response to random terrain, and digital studies of turning stability and
the response to obstacle encounters.
2.1.1 Response to Random Terrain
The design of the vehicle suspension is determined largely by ride quality and
vehicle controllability objectives for const~t-speedtraverse of random
terrain. An analog computer is ideal for studying these characteristics,
since it can readily accommodate random excitation, and the outputs of the
computer can be connected to a mechanical simulator.
The initial analog modeling consisted of a four-degree-of-freedom roll-plane
module with a trilinear suspension and non-linear point-follower wheels with
"scuff" damping. Subsequent analog models consisted of two coupled roll-plane
modules capable of pitch motion a l s o as shorn in Fig. 2-1. A modular approach

1t

1/24
I
 REAR MODULE
ROLL INERTIA

TOTAL MASS 8t
PITCH INERTIA

TORSIONAL STIFFNESS
AND DAMPING

INTERCONNECTING
i
‘ 1 FRONT MODULE
ROLL INERTIA

SCUFF DAMPING
(NONLINEAR) EFFECTIVE WHEEL MASS

WHEEL STIFFNESS (NONLINEAR)

POINT FOLLOWER

‘ RANDOM SURFACE ROUGHNESS

Fig. 2 - 1 MATH MODEL FOR ANALOG STUDY

I/2-2
 to math modelling was selected for the ease with which it allowed coupling of two
or more roll-plane modules for complete vehicle analysis. The major parameters
recorded were chassis accelerations, suspension deflection, wheel load, power
dissipated, and wheel lift-offs. Wheel and suspension characteristics and inter-
modular torsional flexibility were varied.
The trilinear suspension spring used in the analysis combined a linear spring
with a motion-limiting snubber at each end. The three suspensions shown in
Fig. 2 ~ were
2 investigated; these were designated as the 5, 10 and 20 lb/in
suspensions after their linear normal-operating-range spring rates. Static
deflection positions and deflection before snubber contact are also shown.
The nominal stiffness characteristc used for the wheel is shown in Fig. 2-3.
Wheel stiffness ranging from half to twice that shown were considered. The'point
follower assumption which was used was reasonable for the analog studies because
the detailed wheel/terrain geometry characteristics have small effect on ride
qualities. The modelling allowed the wheels to leave the ground when they were
unloaded. The ''scUff" damping mechanism dissipates energy by means of a lateral
scuffing action between the wheel and the lunar surface. This action was
" i modelled as non-linear viscous damping, whose coefficient varied with wheel
"J
1
normal force and vehicle forward velocity. Auxiliary mechanical damping across
i the suspension was also considered.
Four random terrain roughnesses have been defined for the lunar surface in terms
of power spectral densities: smooth m e , rough mare, hummocky upland and rough
upland. Upper and lower bounds for these surfaces are given in Ref. 1. Comparison
of the spectra showed that using rough and smooth mare covered the range, and
these were actually quite close to the remaining two. Therefore, only these
two spectra were used. For each spectrum a straight-line approximation of
the average of the upper and lower bounds in the frequency range of interest
was used. The analytical spectra increased at 6 db/oct with decreasing
frequency down to 0.1 Hz, below which they were made flat to prevent displace-
ment overload of the analog computer, which was scaled for smaller amplitudes
Fig. 2-2 IDEALIZED SUSPENSION STIFFNESSES

VERTICAL DISPLACEMENT, IN.

Fig. 2-3 IDEALIZED WHEEL STIFFNESS

 on the order of wheel and suspension deflections. This low-frequency deviation
would have no significant effect on the dynamic responses since it was well
below vehicle resonances. Uncorrelated random excitations were used for the
left and right sides; left and right rear wheel excitations were the same as
those of the respective front wheels except that a time delay dependent on
vehicle velocity was used.
The initial analog computer work was done using a 2 x 2 roll-plane model.
Results of the suspension variation study using this model are shown in Fig. 2-4.
Astronuat roll and heave acceleration responses are lowest for the softest
'
suspension, as would be expected. Both the 5-and 10-lb/in. suspensions provide
ride qualities within the human tolerance range for both smooth and rough
mare surfaces (Ref. 2). RMS suspension deflections increased with speed. The
time histories of the runs sevelaed that snubber contact occurred infrequently;
on a rough mare at 12 km/hr, only one contact Tn a two-minute run occurred with
the 5-lb/in. suspension, and no contacts were observed with the 10-lb/in.
suspension. (The unexpected order of the deflection curves for smooth mare
operation is attrtbutable to the abrupt non-linearities of snubber contact. )
Also shown are the peak wheel loads that occurred in a two-minute run. Damping
" I
I
power dissipation increased with speed and suspension stiffness; however, a
more marked increase occurred for travel on rough surfaces. Thus, the penalty
for rough mare operation is a bigger drain on the batteries. It can also be
seen that no simultaneous two-wheel lift-offs occurred on smooth mare for the
5 and 10 lb/in. suspension systems. This implies good steering control at all
speeds. Analysis was also made of the peak probability distributions associated
with the suspension deflections and wheel loads for both rough and smooth mare.
The distributions for both parameters were found to be approximately Rayleigh,
as shown in Fig. 2-5 a result which greatly facilitakes determination of
fatigue damage due to vehicle traverse on random terrain. (Determination of
fatigue damage for the wheel is complicated by the fact that the loading cycles
are distributed around the rim rather than applied at a fixed location. This
 W.
a

l I I
-*
0 0 0
m N !2

.
-2 8 -2
,N
.-

W -co

I
m
d P
1
8 0
N
0
8 8 z 0

z
c .
U
I
I .
a
I
Y
z
W
g-
u
Y
$
d s c-
0
8
(0
0
0
d
8
N
0 W
>
n
s
W
U
>
? s n
K
a
N
.-
a
2 z
z
W

m~ N .- 0d
m N 0

W
U
2
I
U
3
0
K
a
m N I 0

1/2-6
 -ANALOG RESULT
- -- RAYLEIGH DISTRIBUTION

WHEEL
LOAD

1Q 2 4 3 4 4 4

A. SMOOTH MARE, VEL = 18 KMlHR

100

80
WHEEL
0 LOAD
z0
60
0
X
u1

0. ROUGH MARE, VEL = 9.3 KMIHR

Fig. 2- 5 PEAK PROBABILITY DISTRIBUTIONS

112-7
difficulty can be surmounted if damage to the wheel for each complete revolution
at a particular load level is estimated and summed for the total number of
wheel revolutions at that particular load level.) The rougher the terrain or
the greater the vehicle velocity, the more the wheel and suspension peak distri-
butions deviate from a Rayleigh distribution because of the non-linearity of
wheel and suspension stiffnesses and wheel lift-off from the ground. The
distributions shown in Fig. 2-5 represent reiatively severe operating conditions;
,
thus deviations from the Rayleigh distribution significantly larger than those
shown would not be expected during normal. operation of the DLRV.
Following the initial 2 x 2 work, the program was expanded to accomodate the
4 x !I vehicle. Results for the 4 x 4 analysis are shown along with comparable
2 x 2 results in Figure 2-6. The intermodular torsional stiffness for these
runs was 200 in-lb/rad, a value approximating that of the actual structure. A
practical range of torsional stiffness was run,' but no significant difference
in the results was observed.
The response obtained with the 4 x 4 analysis are seen to be similar to, but
generally less than, those of the 2 x 2. The slight conservatism of the simpler
analysis is in the right direction. Agreement between the results is close
enough to validate the conclusions drawn from the simpler analysis.
2.1.2 Loads Due to Obstacle Encounter
It is evident that loads due to encounter of large-size obstacles would be
greater than loads produced by response to rough mare traverse. To accurately
evaluate obstacle encounter loads, a more detailed representation of bump/wheel
geometry than that provided by the point-follower assumption was required.
Because such a representation could not be incorporated into the analog analysis
without undue complication, a digital computer analysis was performed. The
digital modelling, which is shown in Fig. 2-7, consisted of a two-mass, two-
degree-of-freedom heave module, with a nonlinear bump-contouring wheel and a
trilinear suspension with viscous damping. Wheel radial loads and suspension
vertical and drag loads were determined for spike bumps of 4, 8, and 12 inches at

1/2-8
 I O
8 5: s? 0

0
8 5: s!

0
8 0 0
8 0 c
8
(D % N
0
(D * V
0

€9 N c 0 m N F 0

0 0
s! 03 % O

€9 cu - 0
L
m N P 0
o
 FORCE

pjg, 2 , F A T H MODEL FOR BUMP ENCOUNTER ANALYSIS

1/2-10
 velocities ranging from 4 km/hr to 16 km/hr. Results are shown in Figs. 2-8 and
-9 for the 10 lb/in suspension and the nominal stiffness wheel. The effects
of other suspension stiffnesses on the loads were small. The limit load of
the nominal stiffness wheel is 300 lbs which corresponds to a 5.5 lunar g
loading. It is seen in Fig. 2-9 that 4-inch bumps, which probably are so
numerous as to be unavoidable, can be negotiated at a l l speeds. Eight-inch
bumps, which would be less numerous and much more visible can be taken at
speeds up to 9 hn/hr.
The effects of varying wheel stiffness on wheel radial and suspension drag loads
'1
"i are shown in Fig. 2-10 for a 10 lb/in. suspension and an 8-in. spike bump.
Three wheel stiffnesses were considered. A s expected, the loads decreased with
decreasing wheel stiffness, but decreasing wheel stiffness implies decreasing
load capability.
3 It might appear, intuitively, that bump-encounter capability could be materially
I
increased by the addikion of a deflection-limiting snubber to the wheel. To
investigate this possibility, a snubber with a spring rate of approximately
1000 lb/in in parallel with the cone wheel was considered. The fourth curve
1 in Fig. 2-10 shows the result. Although this snubber increased the wheel
i
, load capability from 5.5 to 19 g's, it only increased the allowable velocity
over an 8-in. bump from 9 to 12.4 km/hr. Suspension drag loads increased
correspondingly from 140 to 690 lb., which would necessitate redesign of the
suspension arms and chassis support area. It was concluded that the addition
of a wheel snubber could not substantially increase bump encounter velocities
without prohibitive weight penalties. It should be noted that there are no
1
specific obstacle encounter requirements for DLRV; however, it is believed that
the vehicle should at least be capable of withstanding bumps on the order
of 4 in. at maximum speed.

2.1.3 Selection of Wheel and Suspension Characteristics

The information derived from the analog and digital studies was analyzed to
determine the "opthum" wheel and suspension system combination. The analog

I
1 - 1/2-11
 WHEEL OFF GROUND
d
a 250
4) 4 INCH SPIKE BUMP AT 16 KMMR
-I 200 SUSPENSION STIFFNESS = 10 LBIIN.
a NOMINAL WHEEL STIFFNESS
1 50
-I 100
w
w
$: 50
0

-I
a
a
w
>

d
a
4)

z
0
v)
Z
w
h
v)
Dm
v)-I
:-i,,
-75
, , ,
w
u
Y

24

12

0
-12

-24
0 0.4 0.8 1.2 1.6 2.0
TIME (SEC)

Fig. 2-8 TIME HISTORIES OF BUMP RESPONSE

1/2-12
 WHEEL SUSPENSION SUSPENSION
RADIAL 12' DRAG VERTICAL

RESULTS SHOWN
/ : " n FOR NOMINAL
STIFFNESS WHEEL
c
/ AND 10 LB/IN.
SUSPENSION
G
4 8 12 16 4 8 12 16
VELOCITY, KMlHR

Fig 2-9 LOADS DUE TO 4,8, AND 12 IN. SPIKE BUMP ENCOUNTERS

CONE
LOAD = 300 LB
\ (8" SPIKE BUMP, 10 LBIIN. SUSPENSION)

* I
1

VELOCITY. KMlHR
Fig. 2-10 EFFECT OF WHEEL STIFFNESS ON BUMP ENCOUNTER LOADS
9 /

...i
I
I
6

, 3

1/2-13
and digital computer results made readily apparent the effect of varying the
wheel and suspension system stiffness parameters. The stiffest wheels and
suspension springs produced the smallest deflections, but contributed %he largest
wheel loads and vehicle roll and heave accelerations; the softest wheels and
springs produced the smallest loads and accelerations, but the largest static
and dynamic deflections.
Allowable deflections axe limited by vehicle size and obstacle-clearance
considerations. On the basis of dynamic response considerations, optimum
suspension system deflection would be the maximum deflection possible without
snubber eontact. It $hould, however, be noted that occasional contact, which
could be expected with a soft spring, would not have a large effect on rms
acceleration levels and, presumably, on crew comfort.
After examination of the computer data, and evaluation of the tradeoffs, the
combination considered best for both ride qualities on rough terrain and wheel
loads due to bump encounter was the nominal stiffness wheel in combination with
a suspension stiffness of between 7.5 and LO lb/in.
2.1.4 References for Section 2.1
1. "Lunar Terrain and Traverse Data for Lunar Roving Vehicle Design
Study", H. Moore, R. Pike, G. Ulrich. NASA Report to be published.
2. "Preliminary Design Study of a Lunar Local Scientific Survey Module
(LSSM) , Final Technical Report: LSSM Mobility Systems", GM Report
D2-83012-1, July 1966, p. 5-82.
2.2 LOADS AND STRUCTURAL DESIGN CRITERIA
The DLRV system and subsystem components should be designed to the loads and
structural design criteria specified herein. These are based on study and
analysis of the mission objectives and requirements, and previous aerospace
and lunar roving vehicle engineering and development background. Two primary
load environments can be identified:
Stowed lunar transport environment including launch and boost through landing
and deployment
Lunar operation, manned and wunanned

I/2- 14
P

2.2.1 Stowed Condition Loads

The DLRV will be exposed to a variety of static and dynamic loads during its
delivery from earth to the lunar surface. The dynamic environment for Quad I
payload has been specified in a LM interface document. The Quad I V environment
will be similar. LM/Apollo experience has shown that overall design loads for
units Ehe size of DLRV are determined more by dynamic considerations than static
accelerations.

9.- 7
2.2.2 Deployment Loads

'+ i The static loads due to deployment will not be significant for the basic
structure of the DLRV, since the small lunar gravity force will be much less
severe than the forces experienced during delivery to the lunar surface.
Dynamic deployment loads depend primarily on the controlled translational
and rotational velocities of the DLRV at wheel touchdown.
2.2.3 L u n a r Operation Loads
During operation on the lunar surface the DLRV will experience a variety of
ground loads deriving from many sources, such as random surface roughness,
impact with discrete obstacles and braking accelerating, turning, and obstacle
negotiating maneuvers. Loads due to the first two of these conditions came
out of the dynamics studies reported in Section 2.1. Mission design loads
for the DLRV mobility system were established based on rational consideration
of likely loading conditions, in conjunction with the results of the dynamic
* I analysis. A summary of the principal wheel loading conditions, in terms of
.. -1
lunar g units, is given in Table 2-1 while the conditions are discussed in
I more detail in the following paragraphs:
n Z
1
o Nominal static load: This load, which is due to the lunar weight (lgl),
is used as the basis or reference for all the ground load conditions.
The static load on the wheels varies with the mission (e.g., manned
or unmanned, with or without samples) and with the module (power,
control, or science), but for practical considerations only two
nominal loadings were defined: 55 lb for the control module and 40
lb for the power and science modules. These values cover all normal
conditions on level ground; however, they are less than the manned
rescue mode, where degraded operation is assumed.
i

8
H
4 B-
wu '

~ 0

1/2-16
 0 Dynamic operating load: The dynamic studies on the analog computer
utilizing the specified lunar terrain models showed peak loads of
approximately 2 and 4 gl on smooth and rough mres, respectively, at
16 km/hr, during a two-minute run. Allowing for higher peaks during
the longer-tern smooth-mare operation, and selecting a more realistic
rough-mare speed of about 8 km/hr, suggests 3 gl as a reasonable
mission load. This radial load has to be cmbined with loads caused
by maneuvering and braking. An inboard-acting side load due to the
scuffing action of the sMng arm suspension is associated with
vertical loading of the wheel. A frictional coefficient of 0.8
was used to establish the side load shown in the table.
Maximum braking torque: The braking design torque corresponds to a
0
maximum braking effort of 100 ft-lbs .
0 Unsymmetrical braking: This condition assumes maximum braking on two
wheels on the same side of the vehicle; side loads provide the static
balance.
0 Pivoting: This condition involves pivoting the vehicle about a
single locked wheel. Uniform soil pressure is assumed over the
entire footprint length.
0 Turning: This condition corresponds to a sharp turn which puts the
vehicle at the point of insipient instability (all load on the
outside wheels).
0 Maximum radial load: This high-load requirement provides for a number
of radial overload conditions, such as obstacle encounter, longitudional
impact, or a vertical drop.
The dynamic operating loads obtained from the lunar surface spectral densities
do not include impacts with discrete short wavelength obstacles. Separate
digital computer studies reported in the Dynamics Section (281) show %ha% '

a 5.5-g1 capability is required to negotiate bin. spike bumps at 16 km/hr.,

and it is believed that no less capability than this should be provided. The
provision of this high load capability for the science module allows it to
also be used in the manned mode, if called for by future plans.
During unmanned operation, the probability of encountering a large obstacle
is increased, but the driving speeds are much lower. A 5.5 gl capability
allows for longitudinal impact of a vertical wall with a single wheel at a
speed of slightly over 2 km/hr.

1/2-17'
A large free-fall height capability is desirable, since it provides for added
margin during DLRV deployment, and allows for inadvertently driving into a
deep depression. The 5.5 gl load capability provides for a free-fall height
of about 34 inches.
2.3 ENVIRONME3JTAL CRITERIA
The induced and rAtura1 environmental criteria that must be accommodated by
all the DLRV components include prelaunch, launch,'translunar,and lunar surface
conditions. Swmaarized below are the relevant vehicle environmental criteria
as defined in Grummans DLRV final report specifications.
2.3.1 -
(a) Pre-Launch Packaged
-
NOTE: Simultaneous environment and load conditions shall not be considered
unless otherwise noted, at the levels contained herein.
Unless otherwise noted ambient environments are considered to exist outside of
the package.
-
Packaged and Unpackeed the word "packaged" in this specification refers to
containers used for transportation, handling and storage.
Acceleration: 2.67 g ve3tical with 0.4 g lateral, applied simulteneously
to the package. 2.0 g in direction of hoisting (when rings are used,
consider applied to any one or any combination of rings). This condition
applies to all equipnents capable of being hoisted in any direction other
than DLRV axis.
Shock: Shock as in Standard McL-STD-810By Method 516, Procedure 11.
Vibration: The following vibration levels are specified during trans-
portation, handling and storage. Vibration to be applied, along three
mutually perpendicular axes, X, Y, Z to the package. Earth Gravity
Compensation is not required.

1/2-18
 (The: 1/2 Octave per minute, three times per axis from 5 Hz to maximum
Hz and back to 5 Hz).
-
Hz g or D.A.
5-7 2 .5 in. D.A.
7.2-26 * 1.3 g
26-52 .036 in. D.A.
52-500 If: 5.0 g
(4) Pressure: Ground transportation and storage: minimum of 11.78 psia.
- r

Air Transportation: minimum of 3.45 psia for 8 hours (35,000 ft alt).

(5) Temperature:
Ground Transportation: - 6 0 " ~to i-160"~
for 2 weeks
Air Transportation: -45°F to +lb°F for 8 hours
Storage Temperature: -20°F to +llO°F ambient air
2
temperature; plus '360BTU/ft /hr.
up to 6 hrs/day for 3 years.

* !
,
(6) Humidity: In accordance with Standard MIL-S'I!D-81OBy Method 507
except that the maximum test temperature shall be 110°F instead
of 160'~and the minimum test temperature shall be 40°F instead
of 82"~.
-
(7) Rain: As defined in Standard ML-STD-810B, Method 506.
(8) Salt Spray: As specified in Standard MIL-STD-810B, Method 509
(No direct impingement on flight hardware).
( 9 ) Sand and Dust: As defined in Standard MIL-STD-810B, Method 510,
Procedure 1 except that the total exposure shall be limited to
4 hours at a temperature of 73°F and a relative humidity of
22% with the air velocity maintained at 300 If: 200 fpm. (No
direct impingement on flight hardware).
(10) Fungus: In accordance with Standard MIL-STD-810B, Method 508.
-
(11)Ozone: Three years exposure as folluws: 72 hours, at 0.5 PPM,
3 months at 0.25 PPM and remainder at 0.05 PFM concentration.

1/2-19
(12) Hazardous Gases: As defined in Standard MIL-STD-810BY Method 511.
-
(b) Pre-Launch Unpackaged
Ground handling shall not produce critical design loads on the DLRV or DLRV
equipment and shall not increase the weight of the DLRV.
-
NOTE: Simultaneous environment and load conditions shall not be considered,
unless otherwise noted, at the levels contained herein.
Acceleration: 2.67 g Vertical with 0.4 g lateral applied simultaneously,
2.0 g in direction of hoisting in any direction other than the DLRV X
axis direction.
Shock: Shock as in Standard MIL-STD-810BY Method 516, Procedure 1,
Basic Design Test of Figure 516-1. 8

Humidity: Same as pre-launch packaged.

Salt Fog: Same as pre-launch packaged.
Sea-Air Humidity: In lieu of exposure to the individual Humidity and Salt
Fog environments noted above, the following Sea-Air Humidity combined
environment may be substituted. Salt solution per Method 509, of Standard
MIL-STD-810BY except concentration to be 1% instead of 5% by weight of
chemically pure sodium chloride. The chamber to be maintained at
90°F jI 5°F and 85% + 15%/-10% relative humidity for the duration of the
exposure. The 1% salt spray is applied for two-minute periods on an
hourly basis. The duration of the exposure is three da,ys for unsheltered
equipment and one day for sheltered equipment where sheltered is defined
as protection such as by the DLRV skin.
-
Rain: Same as pre-launch packaged but no direct impingement.
Fungus: Same as pre-launch packaged.
Ozone: Same as pre-launch packaged.
Pressure: Ambient ground level pressure.
2
Temperature: -20°F to 110°F ambient air temperature plus 360 BTU/ft /hr
up to 6 hrs/day.
From the time of hypergolic loading of ELM to lift off (T-0):
40" to 110°F SLA cavity external to DLRV
40" to 110°F equipment mounting tray and environment

1/2-20
(11)Hazardous Gases: Gaseous exposure as defined in Standard MIL-STD-~~OB,
Method 511 and MFC Drawing lOMO1071.
(12) Electromagnetic Interference: In accordance with Specifications
LSP-530-001 and MIL-E-605lC.
(13) Sand and Dust: External to the DLRV the particle count shall not exceed
level ~,OoO,OOO of specification Ls~-14-006.
2.3.2 Lunar Environment
Lunar Deployment
Temperature: -+ 3000~
Pressure: 1x mm Hg
Simulated Lunar Dust Basalt particles with density of
15. g/cm3 and grain size of 2mm or less.
Lunar Operation

Pressure 1x mm Hg
Temperature * 3000~
Radiation As defined in NASA's DLRV RFP/RFQ
NO. DCN 1-9-21-00003ANNEX "C"
Electromagnetic Interference In accordance k t h Specifications
LSP-530-001 and MIL-E-6051C
Simulated Lunar Dust Basalt particles with density of
1.5 g/cm 3 and grain size of 2 mm or less.

MOBILITY CONSIDERATIONS
The mobility capabilities of the DLRV depend greatly on its wheel configuration.
As a component, the wheel must satisfy the rigorous DLRV requirements of weak
soil mobility, steep slope and obstacle negotiability, and applied vehicle
loadings. Generally, the larger the footprint area, and the aspect ratio of
this area, the better the mobility performance. Footprint area and its
aspect ratio are functions of both the diameter and the flexibility of the

1/2-21
I
 basic shell of the wheel, hence, of the weight of the shell. Ideally, the
search for the optimum wheel size and flexibility should be based on a
trade-off analysis between wheel weight on the one hand, and a nutriber of
tangible and intangible factors such as power requirements, weight and
dynamic characteristics of the wheel, and overall probability of mission
success on the other hand. Considerations of these factors have been made;
however, the trade-off analyses have been constrained by stowage volume.
The vertical and lateral space allowed by stowage considerations has restricted
the diameter and shape factor of the wheel. Coupled in with these functional
requirements are the constraints of low component weight, high operational
efficiency, and maximwn mechanical reliability.
Weight and reliability received special emphasis; the former because of the
tight weight budget on the DLRV and the fact that six wheels are involved,
and the latter because of its importance in the successful completion of the
prolonged mission.
The cone wheels' strongest points are its light weight, high reliability, and
favorable structural characteristics. Its weight for the large diameter
wheel was the lightest of those compared. Its reliability is enhanced by
self-cleaning tendency and its post-failure behavior. (Cracks at the rim
propagate into lower stress areas where degradation proceeds slowly.) Gradual
stiffening rather than hard bottoming, high torsional and side load capability,
and low one-"g" stress loading (implying long fatigue life) are all desirable
structural characteristics. On the negative side, the cone wheel requires
more stowage volume. With the selected stowage arrangement, much of the
i volume taken by the wheel comes from the crew station, where it would not
.". i) be usable for permanently mounted equipment in any event; however, there is
a loss of about 15% in usable stowage volume with cone wheels. This dis-
advantage is reflected mainly in growth capability, since there is adequate
volume to stow all required equipment and science with any of the wheel
candidates.

If2-22
The conical convoluted wheel offers many of the advantages of the cone wheel
but it takes less stowage volume. The folded geometry produces a moderate
bottoming effect at about 2 g, and it allows for some debris entrapment.
2.4.1 Material Selection
A n outstanding feature of the conical wheel is that it permits fabrication
from a wide selection of candidate materials. Table 2-2 lists these candidates
and compares the resulting wheel designs on the basis of their weight and
structural and mission capabilities. The table presents properties at the
maximwn wheel temperature of 300'F; lower or negative temperatures yeild higher
material allowables and are therefore less critical. The various wheel
designs all have identical spring rates and the same general size and shape.
In the critical weight comparison, the lightest wheel is found to be the aluminum
one, with the fiberglass reinforced plastic (FRP) being a close second. The
other materials are substantially heavier. Overload capability and fatigue life
are also very important. The FRP and titanium wheels, with tlheir low rations
of static stress to endurance limit, should have good fatigue characteristics.
The greater the overload capability of a wheel, the more forgiving it will
be to operational hazards, such as obstacle encounters. The FRP, titanium,
and hybrid wheels (titanium/aluminum) are equally good in load capability;
the aluminum wheel has the least capability. Based on these results, FRP
was selected as the wheel material for DLRV.
Extensive studies of the permanence properties of fiberglass/epoxy show that
in air or under vacuum this material degrades primarily as a result of
W radiation. If the surface is protected from W, there is no degradation,
as shown by aging tests equivalent to one year exposure. The protective
surface selected for this application is a .002-inch pigmented tedlar film
integrally molded onto the wheel. Trevarno F-161 impregnating resin has
been thoroughly tested at Grwrmzan under in-house programs and under a contract
to PZASTEC Corp. to provide data for the new edition of MIL-HDBK-17. This is

1/2-23
rl r l r l r l

1/2-24
 an excellent high-temperature resin and it retains m c h of its strength at
temperatures as high as 400*F. Under low temperature, fiberglass becomes
considerably stronger with little change in its modulus. "E" glass fabric
may be used; however, "S" glass is a better choice since it provides at
least 20% higher mechanical properties, and, in addition, appears to have
superior fatigue properties.
Since the wheel component is critical to the design and performance of the
DLRV, it might be advisable to carry a backup design of titanium or the
hybrid titanium/aluminum into the hardware progrm. In this m y , should any
unforeseen problem arise with the FRP design, a completely compatible alter- '

native would be available. (Grumman is pursuing in-house fabrication technology

for both hot spinning and cold forming titanium.) The hybrid wheel design,
which uses an aluminum hub section and a titanium rim offers good thermal and
structural performance at a relatively low weight.
2.4.2 Wheel Prelim5nary Design
The design criteria for the wheel evolved f r m the configuration studies, and
the mobility, dynamics and loads results discussed previously. These criteria
are summarized in Table 2-3. Because of the large differences in static loads
on the wheels during the mission, it was found advisable to design two wheel
configurations, one for the heavily loaded control module, the other for the
more lightly loaded power and science modules.
'I
The preliminary design of the cone wheel is shown in Figure 2-11. It consists
.._.I of three basic elements, a nominal .060-inch thick conical shell of revolution
measuring 36 inches in diameter and 15 inches in depth, twenty-four grouser
cleats fastened in a space-link arrangement to the wheel rim section, and a
.OgO-inch diameter cable assembly which interconnects the cleats aiding
.
their support. The wheel assembly with cleats attached has a 38-inch diameter
arid a l'l-inch depth.
The conical shell element provides the wheel assembly with its necessary
spring and structural characteristics. A combination of shape, thickness

1/2-25
 TABLE 2-3 WHEEL DESIGN CRITERIA

CONFIG. "A" CONFIG "B" .

Static Design Load 55 40
Wheel Dia. with Cleats (lb) 38 38
Max. Radial Load (lb) 300 220
Max. Lateral Load (lb) 174 127
M a x . Torque (lb-ft) 100 100
Effective Soil Pressure (psi) 1.0 1.0
Thermal Range (OF) k300 +300
Life Cycles (Revs) 106 lo6
Cleat Type Grouser Space Link (Typ.)
Cleat Length (in.) 12 12

Configuration "A" - Control Module Wheels

Configuration "B" - Power and Science Module Wheels

1/2-27
 distribution and materials is used to achieve the desired cone wheel charac-
teristics. The conical shell is fabricated from fiberglass reinforced plastic
using style 778 fiberglass and Trevarno F-161epoxy resin. A thin surface
coating of Tedlar provides protection against ultraviolet radiation and improves
wear resistance. The grouser-type cleat is used to enhance the wheel’s traction
and obstacle negotiation capability. Titaniwn alloy Ti-6AL-4V has been selected
for the cleat material because of its superior resistance to abrasive wear and
its thermal properties. The cleat has an open right-angle cross section
allowing it t o penetrate the lunar soil and perform its grouser thrusting
action. The cleat spacing selected is based on consideration of assembly
weight, cleat reliability, and soil-bridging effects (as reported in Section
2.4.3). Tests have shown that a negligible increase in wheel rolling resistance
is attributable to the cleat system. A feature of the grouser cleat system
‘ 7 is the intercleat tensile attachment which acts to stabilize the cleat in
1 active contact with the lunar surface and distributes the concentrated loads
around the rim. This permits a lighter weight fiberglass cone to be used.
A .090-inch diameter stainless steel flexible cable assembly is used for the
intercleat member. The shape of the wheel is determined by load/deflection
1 requirements and delivery vehicle stowage envelope constraints. The 38-inch
“ J
diameter is the largest stowable wheel. The 17-inch depth and associated
curved shape is designed to provide an initially soft spring rate to develop
to develop a large static deflection and footprint, followed by gently
stiffening characteristics for overload. The hub area is flat and reinforced
“i
.-I with a lightweight core filler to provide additional strength and to
accommodate the wheel drive assembly.

2.4.3 Mobility Perfomance

Vehicle mobility is primarily measured in terms of rolling resistance and
drawbar pull capability. As part of the development program for the cone
wheel concept much test data was obtained at Grwmnan’s lunar test site and
at the Stevens Institute of Technology "Land Locomotion Laboratory" under the
direction of Dr. I. R. Ehrlich. The following discussion will show how mobility
performance is defined and how it is related to wheel and motor gear box
oequirements. Early tests with Gmumnan's 1 "g" LRV simulator and later tests
with prototype 1/6 "g" wheels are presented. The performance data presented
is the latest available and should be considered "representative" of the
anticipated performance of the wheels delivered under this contract. It
should be noted that some of these data require weight ratio scaling to convert
1 "g" earth simulator data to its lunar equivalent.
2.4.4 Rolling Resistance
Total wheel drag R that must be overcome by the wheel motors consists of three
components:
o Rolling resistance RZ due to soil compaction and sinkage z
o Parasitic losses R due to wheel flexure and cleatlsoil scuffing
P
o The gravity component R when climbing a slope
€5
Hence: R = RZ + R + R
P g
The latter component, R is equal to Wsin,< where W is wheel loading a n d d i s
gy
slope angle. R is proportional to wheel loading and could be expressed as qW,
P
where q is generally equal to a few percent. As to the soil compaction resis-
tance, RZ, it can be shown that
n+l
- bPe
RZ - (n+l)k
or
2
- bPe for n = 1 (a)
RZ - 2K
where b is wheel width, pe is effective footprint pressure, k is soil sinkage
modulus and n is the power of the p/z relations, with n = 1 indicating a
linear relation. A one-g version of a 42" flexible cone wheel was tested at
 Stevens Institute and the results are reported in Reference 1. A 1/6-g,
38"-diameter wheel was tested at Gmmrman's Peconic facility at various
loadings and with two types of cleats. The latter wheel simulates the actual
DLRV proposed wheel design. The results of the one-g wheel tests are given in
Table 2-4, The 300-lb wheel loading provides a wheel deflection comparable
to that of the 1/6-g wheel at 55 lbs. The tests on hard ground show a

'"i parasitic loss, R

P'
of 3.8% wheel loading. The test wheel had rubber padded
cleats, however, and the parasitic losses would be less for the rigid cleated
" I DLRV wheel, The effective footprint pressures were calculated in Table 2.4
"i and indicate that about 37% of the measured area is effective in carrying the
wheel load, The measured areas were calculated using full footprint lengths,
i.e., assuming full bridging between cleats,
Two versions of the 1/6-g, 38"-diameter wheel were tested by GAC, one with
7" rectangular wood cleats and the other with.11 5/8" extruded angle cleats.
The deflection characteristics and graphical footprint length of the wheel
are shown in Figure 2-12. The test apparatus consisted of a single wheel
canted at 15" and connected to a frictionless pivot by means of a 12-ft
rigid member, as shown in Fig. 2-13. The wheel was revolved around the pivot
by pulling by hand at a point near the hub. A dynamometer and a recorder
capable of measuring the turning force to the nearest 1/4 lb. was used.
The test was performed on asphalt and soft sand. A roto-tiller was used to
loosen the sand before each pass. Cone index readings of the uncompacted
sand gave a penetration gradient, G, of 1 to 2 psi/in for the top 4". It
is permissible to assme that at these low values, the G parameter is
nearly equal to the soil sinkage modulus k. Additional measurements of
soil constants included the angle of internal friction, @, by means of a
"Sheargraph" and the cohesion, c, by means of a "Torvane". Values of
$ = 35" and c = 0.13 psi were measured.
The 1/6-g wheel test results are plotted in Figure 2-14. The parasitic loss
with the 7" cleats is 2.1% of wheel loading which is appreciably lower
than the 3.8% measured for the one-g wheel.

1/2-30
 TABLE 2-4

TESTS OF 1.g FLEXIBLE WHEEL AT STEZENS INSTITUTE

_- __.____-.-.-- -..I--- - - --
..
_--.

Rubber Padded Rectangular Cleats

Wheel Loading of 330 l b
Wheel Diameter 42 inches
Cleat Length 10 inches = b ----
W r d Surface Fi?m S o i l S o f t Soil

‘-7 1s -
+ -
1
‘1 k, p s i / i n 2

T a r i n g f o r c e , l b . , average 12.4 56.2

of 6 runs

Wheel sinkage, i n , average 2.8 2.1

of 6 runs
‘ 1

1 Measured f o o t p r i n t a r e a , sq. in. 1.53 246

1
Footpring pressure using 2.16 1.34
measured area, p s i
1
3 12.4 43.8
Rolling Resistance, Rz, l b 17.9
e l
I
i
Effective foot r i n t pressure, 5.68 3.62
Pe = (aRZ/b)2
P

I Effective area m t i o , p/pe .38 37

-

1/2-31
 W

s-
RIGID SURFACE

I
a ’

FOOTPRINT LENGTH, 1,IN.

Fig. 2-12 FOOTPRINT LENGTH VS SINKAGE

I 1/2-32
Fig 2-13 TEST APPARATUS FOR ROLLING RESISTANCE

1/2-33
14

TEST DATA
0 7" RECT WOOD CLEATS
0 11 5/8" EXTRUDED ANGLE CLEATS
12 (1 3/8 x 2)

10

20 40 60 80 100

WHEEL LOADING, LB

FIG. 2-14TESTS OF 38" 1/69 WHEEL

1/2-34
The estimated values for the 11 5/8" cleat and k = 1.5 are also compared in
Figure 2-14. These estimates were made using the following procedures:
1. The graphical footprint lengths were determined from Figure 2-12
2. The effective areas were calculated using the above lengths, the
cleat length of 11 5/8, and the effective area ratio of .37 determined
from the Stevens tests
3. The effective footprint pressure, p , and the soil pressure, p = kz
e
are plotted in Fig. 2-15. Curves intersect at the equilibrium point.
Equilibrium pressures are tabulated below for k = 1.5.
w Pe
lb Psi
27.5 0.70
55.0 0.97
82.5 1.22
4. Rolling resistance was calculated from equation (1) using the
equilibrium pressures. A parasitic loss of 2.1% was added.
It is seen in Figure 2-14 that the estimated resistance value is about 80%of
the test value, This is attributed to non-optimum cleat spacing and cleat
deflections on the test wheel. It is assumed that the estimate for k = 1.5
is representative of the rolling resistance for a l l lunar terrain since
(1) cleat design will be refined and (2) the reduction in rolling resistance
of the tracking wheels has been neglected. Grmmman tests of the 38", 1/6-g
wheel with the 7" cleats (k = 1-2) showed a reduction in average resistance to
92% for 2 passes and 87% for 3 passes. Wheel loading, however, was high,
110 lb. Comparable results were obtained from the Steven's tests in soft
soil (k = 1 to 2 ) where the average resistance was reduced to 85% for 2 passes.
In order to evaluate required torque for wheel drive motors, it is necessary
to determine wheel loading for both level and sloping terrain, The results
of these computations are shown in Table 2-5 for representative m a n n e d and

I/2-3 5,
 38" 11% WHEEL

"[
11 5/8" CLEATS

2.0

FIG. 2-lqESTIMATED WHEEL PRESSURES VS SINKAGE

1/2-36
It
 cd
k
8
m
a,
rf

8
m
0
G

"" 1
a

..J
1
P
5

-I
8
a
n W

1/2-37
manned loadings. The total resistance due to the rolling resistance, RZ, and
the parasitic losses, R were taken from Figure 2-14 using the estimated
PY
curve for k = 1.5. Required torques VS. slope are plotted in Figure 2-16.
The average level traverse torque required is 3.9 ft-lb per wheel for
unmanned operation and 5.6 ft-lb per wheel for manned operation. These
were rounded off conservatively to 4.0 and 6.0 ft-lb respectively for all
calculations of level traverse performance.

2.45 Slope Climbing

The slope-climbing ability of a vehicle and its power consumption on level
terrain and on slopes depend on a complex interaction of soil and wheel
properties including sinkage and slip. While analytical estimates are
possible, they must ultimately be verified or supplemented by full-scale
field tests. The pull coefficient, or ratio of draw bar pull to wheel normal
load (P/W), is essentially equal to the tangent of the slope that the vehicle
can climb. The actual climbable slope will be somewhat less than this value
due to such factors as redistribution of wheel loading and reductions of
soil bearing strength due to the slope angle.
Two reasons contributed to the selection of wide cleats for the DLRV wheels.
One reason was to take advantage of the cohesive nature of lunar soils,
and the relatively large contribution of soil cohesion to the total tractive
effort in a 1/6-g field. This factor argues for a large footprint, hence,
for a wide cleat. The selected size, 11 5/8 inches, is the largest that
can be accommodated. The other reason for selecting the wide cleat was to
reduce footprint pressure and, hence, rolling resistance, as shown by test
data in Figure 2-14. Soil bin tests involving a single prototype wheel are
very useful as they permit a better control of the soil and wheel parameters.
DLRV type wheels have been tested at Stevens Institute (Reference l), as
previously mentioned, and at Waterways Experiment Station (WES) (Reference 2).
Both wheels were 42" diameter. A 1-g version was tested at Stevens, a 1/6-g

1/2-38
 50 EARTH WEIGHTS CONTROL UNIT
OR MANNED
-
UNMANNED 972 LB POWER UNIT
MANNED -946L8

40

MANNED CONTROL

$30
I-
ll.
d

z
b
w
z3
3
(r

!j20
SCIENCE UNIT

4
bi P

.- i 10
POWER UNIT

C
I I I I
, 10 20 30 40
TERRAIN SLOPE, DEG

Fig. 2-16 WHEEL TORQUE REOWRED vs SLOPE

t 1/2-39
I
version was tested at WES. The test measurements, soil conditions and cleat
configurations are listed below for the Stevens tests.
SOIL BIN TEST CONDITIONS Ref. (1)
(wheel loading = 330 lbs)
Measurements Soil Conditions Cleat Configuration 11

Draw bar pull vs. slip Cohesionless sand 10-inch flat plate
Torque VS. slip Soft, k or G = l to 2 12-inch, 1* x 1% in.
angle cleat
Sinkage Firm, k or G=8 to 10
Footprint length
Wheel and carriage speed
The draw bar pull, torque and locomotion efficiencies VS. % slip are shown in
Figs. 2-17 to 2-20 and pertinent results are summarized in Table 2-6.
Locomotion efficiency is defined as
7) = pR (1-s)
T
'-",
where R is nominal wheel radius, T is wheel input torque and s is slip. We
may draw the following conclusions from these tests:
o In all four cases, that is regardless of soil or cleat conditions, the
slip on level ground in the self-propelled mode (when DBP = 0) is 1 to
1.5 percent.
o The increase in rolling resistance due to the angle cleats is negligible
on level ground or at low draw bar pull levels,
o The addition of the angle cleats increases the maximum draw bar pull
by 54% in soft soil and by 50% in firm soil. However, a penalty in
rolling resistance, hence, in locomotion efficiency, is paid to
achieve the higher draw bar pull. This penalty is 9% in soft soil
and 22% in firm soil. In conclusion, the angle cleats deliver
significantly more draw bar pull in all soil conditions with less
cost in efficiency in soft soil than in firm s o i l .
The Grumraan wheel achieved a peak pull coefficient of 0.64 in the Waterways
Experimental Station soil bin test, Ref. (2), as against a pull coefficient
of 0.45 in the Stevens Institute soil bin test, Ref. (1). These coefficients
 0

co
0 8 3 0
N
0
8
81 ‘ l l n d

I I I I 1
0 0 0
0
2
0
9
0 x N
0

A3N3131943

1/2-41
 J/
I

I
4w I*
;Y
w

1
I
I

c I I I I
2 W
d x N
0
0

A3N3131333

1/2-43
 0
cu
I
cu

1/2-44
1/2-45
 correspond to maximum climbable slopes of 33" and 24" respectively. The Cone
Index gradient, G, of the soil in both tests was approximately the same;
more specifically, it was 9 psi/in. (Stevens tests) and 12 psi/in. (WES tests).
The discrepancy between the two test results may be adequately accounted for
in terms of the following differences in the test conditions:
o A 1-g wheel under a loading of 330 lbs was used in the Stevens tests,
as against a 1/6-g wheel under a loading of 70 lbs in the WES tests
o A dry cohesionless sand (c = 0) was used in the Stevens tests, as
against a moist, relatively finer sand with a measured cohesion
of 0.16 psi in the WES tests
Figure 2-21 shows the resolution of these data and the theoretical relationship
between maximum climbable slope (pull coefficient) and wheel loading at
various soil cohesions. Steven's P/W of .45 for c = o is considered
independent of wheel Loading. The effective footprint length necessary to
account for the WES results at c = .16 was used on conjunction with
Figure 2-22 to determine the assumed footprint length, le, VS. loading.
P/W was then computed from
." i p/w = (P/NC = + (P/W
= (P/Wlc = 0 + le bc/W
= 0.45 + 12 le c/W
where b = 12 in. = cleat width
Cohesion in lunar soils is necessary if the vehicle is to climb a 35" slope
in the low gravity of the moon. Notice how, at a given cohesion, climbable
slope increases with decreasing wheel loading
Although the wheels tested are not identical to the DLRV wheels, the results
are representative of DLRV slope climbing capability. These results indicate
that the DLRV wheel can climb a 35" slope under expected wheel loadings if the
a soil has a cohesion of the order of 0.1 to 0.2 psi. It is recommended that
e 4 additional tests be performed to optimize the cleats at wheel loadings from
10 to 50 lb in soils with cohesion from 0.1 t o 0.4 psi.
Although these data had been presented in the DLRV final report, insufficient
time was available to incorporate the latest cleat configuration test presented
herein into the overall wheel design shown for the DLRV. Section 3 of this
report relates how these data were used to design the cleat for this study.

1/9-46
 1 /-
0 I I I I I
20 40 60 80 100

W, LB

1.0 -
0.84 -
c = 0.2 \-
.
3 0.7 - REQ'D LIMIT

n
sLU 0.68 - 30
WES TEST

0
U
U
w
x
n \
0 w- c = 0 (STEVENS TESTS)
0 0.45 0
-f
3 '3
-
v)
3
a. 0.36

REAR.WHEEL, EITHER MODE

0.18' \
\
MIDDLE WHEEL\ \
0 I \ I , I \ I i I
20 40 60 80 100

WHEEL LOADING, W-LB

Fig. 2-21 MAXIMUM SLOPE CLIMBING ABILITY, 42" WHEEL, 12" ANGLE CLEATS

1/2-47
 2.4.6 References for Section 2.4
1. Stevens Institute of Technology, Davidson Laboratory, "Tests of Lunar Rover
Wheel", by L. I. Leviticus and I. R. Ehrlich, Report 1429, November 1969.
2. U, S . Waterways Experiment Station, "Study of Mobility Performance and
Slope-Climbing/Traversing Ability of Lightly Loaded Wheeled Vehicle on
Soft Soil", Fourth Monthly Progress Report, 1-30 Sept. 1969, Contract ?

DPR H-58504A.
2.5 SUMMARY OF DYNAMIC DESIGN CRITERIA
The following data were extracted from Gmunmans DLRV final report specifications,
updated to reflect additional testing and revision to the cleat configuration.
.-
I

*j The wheels shall be conical in shape with grouser-type cleats attached to the
rim-running surface. The wheel material shall be fiberglass reinforced
plastic with eight plies at the hub and rim and five plies in between. .The
hub section shall be reinforced and configured to accept the center line mounting
of the wheel drive assembly. The wheel assemblies shall be designed to
satisfy the following operational requirements:
Control Module Wheels
Static design load (lb) 55
Max. radial load (lb) 300
Max. lateral load (lb) 174
M a x . torque (lb-ft) 100
Static lg deflection (in) 1.8
Drawbar pull per unit
weight (minimum) .6
Nominal spring rate about
1 g static (lb/in) 60
Weight (lb maximum) 15.0
Wheel diameter (in) 38
Cleats shall be formed 9f 0.025 inch titanium. Each wheel shall have
30 cleats equally spaced about the rim.

1/2-48
 3 DESIGN DEVELOPMENT AND SELECTION

3.0 GEPIERAL
The primary study objective was t o design and deliver to NASA three test wheels
compatible with the DLRV mission requirements. An iterative design evaluation
was followed based on Gmwrman's previous experience with its f u l l scale DLRV
simulator, design studies conducted for the DLRV phase 3 study, and additional
emperical test data obtained during this study.

3.1 STRUCTURAL ANALYSIS

The cone wheel structure is a thin open shell of revolution consisting of a
r -
hub, a conic section, and a rim joined by two transition radii. The rim
along with the First transition radius serves as a hoop to distribute the load
to the conic section and a l s o prevents the conic section from buckling. The
conic section carries the load to the hub as a large thin-walled tapered tube
and also acts as a cantilever beam for the lower segment of the conic section.
In designing this type of wheel the objective is t o keep the stresses in the
rim area about equal t o those in the hub area; this concept of having
balanced stresses enhances the overall design.
Wheel deflections at the rim can be very large, making any analysis of the
area quite difficult. Experimental test results have been used instead t o
form an emperical basis for the cone wheel design.
The area from the hub to a region near the rim is not subjected to such large
deflections, and can be more readily analyzed. Grwnman has a shell computer
program (STARS 11) that is capable of analyzing any structure that can be
idealized as a combination of varying thin surfaces of revolution, including
cylinder, ellipsoid, ogive, parabiloid and cone. The program obtains solutions
to the equations of elasticity using assumptions of small displacement theory,
and can handle symmetric and unsymmetric loadings. It has been used successfully
at Grwaman on a variety of shell structures including, among other, LM propulsion
tanks, LM landing gear and foot pads. This program was used to size the cone
wheel in the l o w deflection regions away from the rim.
Extensive full scale testing prior to this study on a 1/6-g FRP cone wheel,
and 1 "g" earth equivalent were used to augment the analysis. Figure 3-1
shows the earlier instrumented wheel undergoing static testing. Normal
loading on a flat surface, concentrated loading, and combined normal and
drag loading conditions are shown. Note that the deformed shape under the
concentrated load differs little from the flat surface load, indicating
that there is no extreme stress build-up in the vicinity of a concentrated
load with a cone wheel. The detailed results and discussion of this prior
testing is included here to show continuity, and document under one cover,
all relevant test data.
3.2 PRIOR G R W TESTING

3.2.1 Strain Gage Instrumented Test Wheel

Figure 3-2 shows the details of the 1/6 g cone wheel test article. The test
.
wheel was made of 181 volan "A" fiberglass reinforced plastic (FRP) Ten
strain gages were provided, covering the critical circumferential and
meridional locations of interest, adjacent to the wheel rim. High elongation
- - - -
strain gages, identified by EP 80 250 MM 120, were used. The hub
area was not instrumented, because it was decided to concentrate the
instrumentation in the rim area of the wheel.

3.2.2 Test Set-UQ

A simple test fixture was erected to obtain load-deflection characteristics
of the cone wheel. The test fixture, shown in Figs. 3-3 and 3-4 was designed
to rigidly restrain the wheel at a 15" positive camber angle.* The test
wheel was attached at the axle to a rigid hu5 assembly, similar to the Grwnman
simulator installation. Deflection at the wheel was accomplished by vertically
raising the contact surface platform via two 2 calibrated hydraulic actuators.
This platform was supported and guided to eliminate bending and local
deformation,

* This camber angle minimizes the scuffing component of the footprint of the
deflected wheel.
 (A) TEST WHEEL AT STATIC LOAD

(6)STRAIN GAGED WHEEL WITH POINT LOAD

(C) STRAIN GAGED WHEEL WITH NORMAL AND DRAG LOADS APPLIED
Fig. 3-1 INSTRUMENTED DLRV WHEEL UNDERGOING TESTS
I /1-4
' 7
i
6 1

I
*.. J

.I
._ i

113-5
 Figure 3-4

’ “I

1 ”g” .Cleated Test Wheel Under Load

T /1-6
 Variation of friction coefficients were tested by providing a wheel surface
contact area, consisting of a) rubber/t/l=0.8, b) t e f l o n e 0.1 and c) roller
supported intermediate board4/ = 0.0.
Provisions were made in the test fixture to allow rotation of the test wheel
in 11y increments in order to obtain circumferential stress variation.
The combined vertical and drag load condition was simulated by applying a drag
load to the roller supported board, via a calibrated spring scale at a given
-:-I vertical load as shown in Figure 3-5.
I
Step obstacle characteristics were evaluated with the wheel loaded against
a 90" V-Block as shown in Fig. 3-6.
For purposes of comparison and reference the lg wheel was also load-stroke
tested and the change of curvature recorded. The deflected wheel is shown
in Figure 3-7. This wheel was not strain gage instrumented.
3.2.3 Test Results
The main objectives were to obtain the vertical flexibility characteristics
which are plotted in Fig. 3-8 for the lunar wheel and Fig. 3-9 for the earth
wheel, and stress distributions which are presented in Figs. 3-10 thru 3-18.

Graphs in Figs. 3-10 thru 3-13 present the circumferential stress variation
starting at 0.5 in. from the edge to 11 in. along the wheel meridian. Figs.
3-14 thru 3-18 give the meridional stress distribution at varying positions
on the circumference, under different loading conditions. These graphs are
otherwise self-explanatory.
E l
A vertical load of up to 250 lb. was applied to the 1/6 g test wheel and a
I. I
maximum stress of up to 24000 psi recorded. No permanent set was observed
at aay of the tested conditions. The lg wheel was loaded up to 1200 lbs.
also without permanent deformation.
The plotted stresses are based on a conservatively assumed Youngs modulus of
6
elasticity E = 3.3 x 10 psi for the 181 FRP material.

I/3-7
r
i
l
FfGURl7 3-5

- *1 i

w
cd
k
R
a
c
cd
 FIGURF: 3-6

I/3-9
 FIGURE 3-7

1/3-10
 1/6 "g" FRP CLEATLESS WHEEL
VERTICAL LOAD VS STROKE

FIGURE 3-8

1/3-11
 1 "g" FRP CLEATLESS WHEEL
VERTICAL LOAD VS STROKE

FIGURE 3-9

. ._J

1/3-12
 STRESS DISTRIBUTION TESTS
1/6 “ g ” FRP CLEATLESS WHEEL
I
,
I
VERTICAL LOADING ON TEFLON 4= .1
j *
GAGE 0.5 I N , FROM RIM EDGE

FIGURE 3-10
 STRESS DISTRIBUTION TESTS
i
i 1/6 "g" FRP CLEATLESS WKEEL

VERTICAL LOADING ON TEFLON /y=.1

GAGE 2.5 I N . FROM RIM EDGE

FIGURE 3-11

1/3-14
 STRFSS DISTRIBUTION TESTS
1/6 "g" FRP CLEATLESS WHEEL
VERTICAL LOADING ON T E F L O N q = .1
GAGE 5.0 I N . FROM RIM EDGE

FIGURE 3-I2

* -

*I
i

1/3-15
 STRESS DISTRIBUTION TESTS
1/6 "g" FRP CLEATLESS WHEEL
VERTICAL LOADING ON TEFLON/cI = .1
GAGE 8.0 I N . FROM RIM EDGE
(11.0 I N . FROM EDGE SIMILAR)

FIGURE 3-13

1
i

I
i
- 3

\- I

1/3-16
 STF33SS DISTRIBUTION TESTS
1/6 FRP CLEATLESS WHEEL
i "g"
VERTICAL LOADING ON R U B B E R A = .8
= 180"

FIGURE 3-14

i
3
 STRESS DISTRIBUTION TESTS
1/6 "g" FRP CLEATLEXS WHEEL
VERTICAL LOADING ON RUBBER&= .8
o(=3-24"

1/3-18
1/3-19
 STRESS DISTRIBUTION TESTS
1/6 "g" WHEEL W I T H 1 "g" CLFATS
VERTICAL LOADING ON TEFLON&= 0.1
= 180"

FIGURE 3-17

1/3-20
 STRESS DISTRIBUTION TESTS
1/6 "g" WHEEL WITH 1 "g" CLFATS
VERTICAL LOADING ON TEFLON& = 0.1
= 112"

FIGURE 3-18

1/3-21
3.2.4 Evaluation of Results
0, The overall relative performance of the scaled down 1/6 g test-wheel is in
good agreement with the lg wheel. The static footprint of the cleated
1/6 g wheel exceeded that of the lg wheel.
0" Extrapolated strain gage data indicate better than 300 lb. radial load +

capability for the 1/6 g test wheel,

0 The test wheel was capable of transmitting all the preliminary ground
loading conditions then defined for the DLRV without permanent deformation.
0. Two maximum stress areas were indicated:
Hoop stresses at the rim edge, and
Meridional stresses at the transition between the cylindrical and conical
surface of the wheel. The provision of eleats effects an increased
stress level at (a), and a stress reduction in (b). The most critically
stressed area, both for ultimate and repeated loading, occurs at the
rim edge of the wheel, where the following stress levels were recorded:
(1) Nominal static loading 50 lb. f C10,OOO psi
(2) Vertical V = 150 lbs. with high drag
load D = 110 lb. f e 18,000 psi
(3) Radial overload 250 lb. f e 24-25,000 psi
0. Although these tests were not conducted to failure, the results indicate
that the failure mode will consist-ofa meridional crack initiated at the
rim edge. Such a failure is not envisioned to be catastrophic to the
wheel, because the decreasing stress level along the conical surface,
will retard crack propagation.

3.2.5 Conclusions
The relatively low 1 "g" operating stress level (less than 10,000 psi) compared
to the allowable fiberglass flexural strength of about 59,000 psi indicates
6
ample fatigue life (10 cycles) for the tested wheel. This is particularly
BO in view of the lack of stress concentrations inherent in the cone wheel

design.
 3.3 &EL DESIGN "A"
G m a n ' s DLRV phase "B" final report defined a cone wheel configuration based
on the ptructural analysis and testing discussed in paragraph 3,2, and the
structural and mobility design criteria discussed in section 2.0. Because of
the substantial effort applied t o the DLRV design, that configuration (Fig. 3-19)
was selected as test wheel "A" for this study. The cleat configuration was
revised (Fig. 3-20) to reflect the improved mobility performance noted in the
test data presented in Section 2.4.
Twenty four-twelve (12) inch long aluminum angle extrusion cleats were provided.
The cleats were mechanically fastened to the rim at two points, and were inter-
connected to each other by a cable restraint system.

3.3.1 Fabrication Process - Test Wheel "A"

The test wheel was manufactured of 181 fiberglass cloth and epoxy resin using a
1 male plaster mold and structural wet layup techniques. The nuber and orientation
of the plies is shown in Fig. 3-19. An average wheel gage of .060 was provided
I exclusive of a single-layer rim reinforcement. The low-temperatwe resin
I
system pemited quick, low-cost plaster mold changes for evaluating configuration
i
1 changes. Differences in modulus of elasticity between the test wheels and the
."I
final wheels fabricated later with a higher temperature resin system are
negligiable .
3.3.2 Test Plan
The wheel was instrumented to measure its load-stroke characteristics conincident
with the stress distribution within the wheel.
Test wheel "A" had 16 high-elongation strain gages (EP-08-250-120) installed to
measure hoop and axial stresses from the rim to the hub. Figure 3-21 shows
the exact gage locations and measurement axis. In order to obtain radial stress
distributions the test wheel was incrementally rotated and the load-stroke test
repeated.

I 1/3-23
 FIGURE 3-20

I N I T I A L CLEAT CONFIGURATION

1/3-25
 FIGURE 3-21

INSTRUMENTED TEST WHEELS

I
I
JI
 The wheel test fixture was designed to rigidly restrain the wheel at a 15"
positive camber angle to the contacting surface. Wheel deflection was accom-
plished by vertically raising the contact surface. The contact surface was
teflon coated and free t o scuff laterally providing a m i n i m friction
coefficient interface. The test fixture and procedures were identical to those
noted in the previous tests (para. 3.2).

3.3.3 Test Results

Analysis of the wheel test data with no cleats attached indicated good footprint
characteristics at the nominal 1-g lunar load (55 lb), and a projected capacity
to easily support a 300 lb load with the cleated wheel. A maximum stress of
25,000 psi at 250 lbs load and 10,000 psi at 110 lbs were measured at the wheel
hub. This stress level would confidently assure no fatigue failures for
DLRV mission.
The addition of cleats' stiffened the wheel spring rate and as anticipated
increased its load carrying capacity over the basic wheel. Figure 3-22 shows
both load-deflection curves. However, rim distortion, attributable to the
cleats, was noted at the higher loadings, and the test was terminated prior
to obtaining the stress distribution data. During the test, the cleats
were noted t o be in full ground contact from 55 t o 300 lbs.
Figure 3-23 a and b shows how the load applied to the leg of the angle is
' * j
1
reacted on the relatively soft rim outside of the base of the bolt pattern.
L . .I
This causes a moment t o be applied to the rim, resulting in a local deflection
1
* I with sinsoidal characteristics. Figure 3-23 shows a revised configuration
'IT" cleat with the applied load passing inside the bolt pattern, resulting in
no moment applied to the rim.
The "T" section cleats were incorporated on wheel "A" and the tests were
satisfactorily completed. Measured stress distributions with "T" section
cleats are tabulated in Table 3-1. Figures 3-24 and 3-25 summarize in plotted
form typical meridional and hoop stress distribution. Maximum stress values
are shown in Figures 3-26 and 3-27. These stress levels confidently indicate

1/3-27
i
 (a) Sketch shodng sinsoidal r i m deflection a t 300 lb. load.

(b) Moment applied t o r i m i s approximately equal t o P x e

b? -_ e- (4
: \

(c) Since p i s i n s i a e t h e b o l t pattern there i s no monemnt applied t o r i m .

1/3-29

k
r i
<'"- 1
si

1
""4

L-. J
-* i
I
\

1/3-33
 I

adequate fatigue life for the DLRV missions, coincident with a wheel that provides
the required spring rate and footprint characteristics necessary for operating
on real lunar soils.
3.4 DESIGN ITERATION
The design of a second test wheel (wheel "B") was initiated to further improve
the weak soil mobility while maintaining the 300 pound load capacity. Figure
3-28 shows the revised wheel configuration. The diameter was reduced from
36 to 34 inches while the wheel depth was increased from 15 to 16 inches. me
reduction in diameter was done in order to allow a larger transition radius
(9.7" radius) between the conic section and the rim. The larger radius was
intended to provide a lower initial spring rate, which would give a larger
static footprint. Both modifications are in the direction of increasing the
static ground contact area for improved traction.
Based on the problem noted with the 90" extrusion cleat, the "T"-section cleat
of the same dimensions WAS used for this wheel.
No changes were made in the manufacturing materials techniques, or instmentation.
Wheel gage was also maintained, except in the rim to conic section transition
area where the gage was reduced from 0,070 to 0.060.

3.4.1 Test Results

Wheel "B" was tested using the same loading fixture and test procedures.
Analysis of the data indicated the anticipated general softening of the
spring-rate on wheel "B" was not realized (Fig. 3-22), indicating that the
deflection is more strongly influenced by wheel diameter than the increased
(7" - 10") rim to conic section transition radius, the increased wheel depth
-
(15" 1-6")~or the reduction in gage (.070" - .o6ot1) at the rim transition
area. The rkn/transition area acts as a hoop which must deform in order to have
-
any wheel deflection. The reduced diameter hoop (36'' 34") increased the
wheels stiffness, negating the expected effects of the other changes.
Additionally noted was a local deflection on the "B" wh
cleats bearing on the lower, transition radius between
causing the initiation of a secondary buckle at loads akfove 150 pounds. A
combination of factors including a decrease in the rim-cone transition
-
gage, the increased rim conic section transition radius, and the unyielding
"T" section cleat contributed to the problem. Normally, when the wheel is
stroked without cleats the lip curls upward from the ground at highloads,
enhancing the local strength. When the rigid cleats are installed, this
curling is completely inhibited on a rigid surface. This restraint of the
rim by the cleat causes a local buckling condition. The larger transistion
radius of wheel "B" is less stiff and therefore buckles more readily.

Close examination of wheel "A" showed a similar problem, however, only slight
local deflections occur at about 250 pounds and at the 300 pound design
point a much less sever condition exists than that noted on whee1"B".
Neither wheel exhibited any problems in the uncleated configuration, and both
used the "T" cleats for all other tests. The stress distributions shown
in Table 3-2 do not reflect this condition because of it's localized nature.
Figures 3-29 and 30 and 3-31 and 32 show graphically a typical and maximm
stress profile.

3.5 FINAL DESIGN SELECTION

Based on these data wheel "A" with a revised cleat configuration, was selected
as the final design. The revised cleat is designed to deflect under eccentric
loads in excess of two lunar 'lg's'', thus minimizing the adverse load transfer
to the wheel, while still providing the aggressive soil-wheel interface necessary
for tractive mobility. A prototype titanium deflecting cleat was tested on
wheel "A" in the load-deflection fixture prior to final fabrication.
Figures 3-33 and 3-34 show the nominal cleat footprint, and the cleat reaction
to an eccentrically applied load inboard of the rim edge. No local derlections
were noted in the wheel, and the cleat returned to its original configuration
when the load was removed. Since previous instmented tests had not detected
high stress concentrations no additional quantitative data was obtained with

1/3-37
m
m

,
m
H
LW - 1rV- 110- IOU- 13v-
160' 170" I goo 200° 210"

140'
220'

130
230

120
240

1i o
250

Io(1
260

9c
27C

8c
28c

7(
2%

6(
30(

5(
31f

4
32(

70
1/3-43
 FICT'URE 3-34

Fr
0

ki

1/3-44
the deflecting cleat. Tests of wheels "A" and "B" used a 24 cleat configuration.
A 30 cleat configuration was selected for the final design. This increase will
enhance the tractive performance of the wheel by reducing ground pressures,
plus providing an increased area of rim coverage minimizing the possibility of
impact damage to the wheel. Figures 3-35 and 3-36 show the selected wheel-
cleat configuration and assembly.

3.5.1 Final Design Characteristics

This study did not require testing of the final design, however, based on wheels
"A" and "B" the following characteristics are anticipated.
3.5.1.1 Spring Rate
The spring rate of the uncleated wheel should closely approximate that of
wheel "A", being affected only by a small difference in the modulas of
elasticity ("E") between the wet layup and the $repre# Trevarno 161 fiberglass
systems, however, the cleated wheel may exhibit a slightly lower spring rate due
to a less restrictive cable system with the deflecting cleat.
3.5.1.2 Stress Distribution
The measured stress distributions of wheel "A" should not be significantly
different than that of the delivered wheels. Slight changes could be
anticipated primarily in the rim area, again due to the revised load transfer
behavior of the deflecting cleat.

1/3-45
 4 FABRICATION PROCESS

4.0 FABRICATION REQUIREMEN'llS

A Total of five wheels, two for configuration development testing, and three
for delivery to NASA for DLRV qualification testing were provided. Both the
test wheels and the deliverable wheels will have grouser cleats as developed
during this study.

4.1 TEST WHEELS

The major difference between the development test, and delivery wheels is
in the tooling and laminating processes used to fabricate the fiberglass
shell. The test wheels used inexpensive short lead time, easily modified
plaster tooling and structural wet layup techniques. The basic wheel was made
of 181 Volan "A" fiberglass reinforced plastic (FRP) withEpon 828 resin and
Shell curing agent A. During layup the plies were checked for fiber orientation
and seam location on the mold. After completion of the layups a vacuum was
drawn around the tool removing any entrapped air o r excess resin. The wheel
was then oven cured at 200°F for 2 hours.

4.2 DELIVERABLE:WHEELS
The final three wheels which were delivered to NASA were made from a pre-
impregnated fiberglass (style 7781) cloth containing Trevarno ~161 epoxy resin.
The individual plies were laid up as noted on the drawings, and a vacuum bag
applied as before to remove any entrapped air or resin. The wheel was then
autoclave cured at 50 psi for one hour at 350'F. A steel male mold was
fabricated to withstand the required cure cycle without degrading. The Trevarno
was selected because of its ability to withstand the extreme temperature range
associated with lunar day and night operations.
The bare wheel was coated inside and outside with a TFE filled polyurethane
paint to eliminate any ultra violet (UV) induced degradation, and to protect
the wheel from light scuffing.
4.3 CLEAT FABRICATION
Cleats for the test wheels were fabricated from standard aluminum extrusions.
The deliverable wheels were fitted with cleats to satisfy the wear and loading
requirements associated with the DLRV mission. The cleats were made of 0.025
gage 6AL-4V titanium, with standard AN3C-4 hardware used for attaching them to
the wheel. RW88 potting compound was used between the raying surfaces, and
also as a fillet seat to minimize the possibility of debris ingestion between
the wheel and cleat. A total of 30 cleats per wheel are provided with an
interconnecting 1/16 inch stainless cable. The cable system stabilizes the cleat
at the wheel soil interface, and helps to more evenly distribute both torque
loads and any higher than nominal vertical loadings.
4.4 WHEEL WEIGHT
The total weight of the completed wheel assembly is approximately 14 lb. 9 oz.
The basic shell with the polyurethane coating weighs 9 pounds, with the cleats,
cable assembly, AN hardware and potting compound accounting for the remaining
5 lbs. 9 OZ. Individual wheel weights are listed below.
Wheel # Weight
1 14 l b s 11 oz
2 14 lbs 8 oz
3 14 lbs 9 oz
Weight reductions on the order of 0.5 lbs per wheel can be achieved with the
use of lightening holes in non structural areas of the cleat, and by replacing
the AN3C-4 hardware with lighter weight fasteners with a shorter shank.

I/4-2