Document
THE EVOLUTION OF THE VIKING LANDING GEAR a JOE C. POHLENb BR ueE D. MAYTUMc IRA W. RAMSEyd U. J. BLANCHARDe ABSTRACT The primal-y function of the landing gear is to dissipate touchdown energy in a controlled fashion that minimizes the landing shock for onboard components while maximizing ground clearance and the probability of land ing stability ove:>: the range of possible surface and touchdown parameter vari- ations. Many other rnission requirements and constraints were translated into the evolving Viking landing gear design as they arose, Hence, design considerations included such factors as prelaunch heat sterilization and non- contaID.ination of the Martian landing site, gear stowage and deployment, terxninal descent engine shutdown initiation, structural load attenuation, ha.rdl soft .landing surface capability, reliabilitYi weight, and post-landed stability.
The landing gear, which was selected on the basis of proposal-phase trade studies, consisted of three inverted·-tripod legs with crushable honey- co'!ub elements in the main struts and omnidirectional crushable footpads, both for energy absorption and strut load attenuation. The gear design evolved through several interrn.ec1iate configurations during early analytical studies and development test programs as the functional specifications for the landing gear were definitized. The final flight-design landing gear consists of three inverted-tripod landing legs, with optimized crushable honeycomb elements having return-stroke load ca.pability in the main struts, deformable load limiters for the attachrnent of the split bipod struts to the lander body, and a footpad having both hard surface capability and soft surface bearing load en- hancement.
aThe work reported here was perforrned under contract NASl-9000 bUnit Head, Martin Marietta Corporation-Denver Division cSenior Engineer, Martin Marietta Corporation-Denver Division dEngineering Mechanics Manager, Viking Project Office, NASA- Langley Research Center eResearch Engineer, Impacting Structures Group, NASA-Langley Research Centel' 218 JPL Technical Memorandum 33 -777 INTRODUCTION The configuration of the Viking landing gear at the Hril.e of contra'ct award in May 1969 was conceptually silnilar to the Surveyor gear. .A crush- y able/shearable footpa.d was fastened to the end of the tripod. The original footpad is shown in Figure 1. The main strut for Viking utilized crushable honeycomb with multiple force levels instead of the pres surized oleo strut of Surveyor. Component footpad, attenuator and limiter development tests, vehicle stability analyses, and sub-scale model tests were conducted through 1970, 171, and '72 with full scale vehicle structural model drop tests in 1973 ll and three system-level vehicle drop tests on a "flight type lander in 1974.
The flight gear, following one of these drop tests, is pictured in Figure 2.
The requirements which were imposed on the landing system during the course of the program are listed in Table 1.
Table 1. Viking Landing Gear Design Criteria Parameter Contract Go. Ahead As Flown Vertical Velocity 3 ± 1. 5 m/sec 2.44 ± .9 m/sec (10 ± 5 fps) (8±3fps) . Horizontal Velocity ± 1. 8 m/sec ± 1.22 m/sec ( ± 6 fps) ( ± 4 fps) Engine Cut-Off Altitude 3±1.5m at touchdown (10 ± 5 feet) Engine Cut-Off Sensor by Radar switches in legs Stability all 19° slopes 99. 7% stable Package Loads 80 gls 30 g's Clearance 22 cm 22 cm ( 8.66 in. ) 8.66 in. ) of Friction 1.0 for stability 1. 0 for stability Coeff.
O. 2 for clearance O. 2 for clearance JPL Technical Memorandum 33-777 219 } FOOTPAD/LOAD LIMITER DESIGN The prim.ary function of the crushable footpad em.ployed in the original gear design (circa 1969) was to limit the lateral loads applied to the land~r.
This design concept relied on the fact that the shear strength of the pad honey-· cOlub is less than the longitudinal crush strength. The footpad was fastened to the inverted tripodal gear by means of a IIU" joint to prevent bending moments in the main strut and the sec(;>ndary bipods, in the event that the pad hit an uneven surface. A footpad rollover problem was uncovered early (1970) in testing conducted at Langley Research Center. The problem shoym in the photo of Figure 1, was especially evident when landing on a hard, high" friction surface. These tests showed that the crushing fcrce dropped drq.s- tically for loads applied at angles to the honeycomb cell axes, so tkq; the omnidirectional capability of the crushable honeycomb footpad was very poor and unpredictable. Also, sharp or protruding immovable rocks easily pene- trated the honeycomb while providing little or no energy absorption. These results led to the deletion of the crushable rotating footpad and prompted the development of a fixed conical footpad design. A solid spheri.cal nosecap was selected to keep high load, hard surface landings from "dumping" high moments into the bipods, The soil penE;tration tests pictured in Figure 3 dictated the addition of an integral revers~ flange skirt on the cone. This thin flange could not carry point loads, and therefore could not cause ex- cessive moments in the bipods. The impact velocity for all the tests shown in Figure 3 was 2.5 m/sec. The soil density was 1. 4 g/cm for a and b (sand), 1.45 g/cm for ~ and d (lunar nominal), and 1.6 g/cm for ~ ~nd
i. (also lunar nominal). Drops pictured in a, c, and e were conducted with a
bare conical pad, while E., ~, and £ had the reverse flange. The final flight
des ign is shown in Figure 2.
At the time that the conical pad was implemented, the secondary struts (bipods) were conceived to be telescoping members, permitting both tension and compression stroke and having load limiting capability while carrying significant moment and axial loads. Another haunting requirement, that of post-landed stability, stood a cha.nce of loosing ground. While yet in the paper phase of the above nightmare, the project "invented" the idea of fasten- ing a load limiter to the body of the lander. A very simple H-beam type bipod could be fastened to this and when the axial load in the bipod member exceeded the bending capability of the limiter, it "stroked. A significant analytical and development test program took place to get the actual limiter performance characteristics now employed on the lander. At about the time the initial limiter hardware was being delivered, significant site alteration test results allowed the lander to be flown all the way to the surface. This change reduced the maximum touchdown kinetic energy by a factor of four.
As a result, the load limiter design went from the one pictured at the top of Figure 4, which weighed 1. 5 kg , to that of the other three limiters shown in the same figure with a weight of 0.45 kg. The limiter on the right is shown 220 J"PL Technical Me.morandu.m 33-777 as-built with the center one tested and exhibiting approximately 2 cm of stroke and the lim.iter on the left has been stroked 6.0 cm. The flight configuration lim.iter design fabricated from fully armealed stainless steel had load require- ments shown in Figure 5, with typical test performance as shown.
MAIN STR UT ATTENUATOR DESIGN The primary strut transmits the majority of the vertical load to the lander body, and hence, the landing deceleration applied to on- board com- ponents is controlled by the selection of attenuator crushing force levels.
More importantly, the main strut attenuator must dissipate most of the touch- down energy, particularly in the event that the footpad can slide out across the landing surface with little frictional resistance. The force levels carried in each of the strut members, and the changing gear geometry during landing determine the stabilizing {and destabilizing} moments which act about the lander center of gravity.
Crushable aluminum honeycomb (H/C) tubecore was selected at the outset to provide the necessary landing energy absorption for several reasons: the desired crushing force levels can be obtained by varying the core cross- sectional area and the skin thicknesses; the core is very lightweight; the core can be stroked over 80% of its initial length before bottoming; and the crush- ing behavior is unaffected by long- term exposure to the interplanetary thermal/vacuum environment. These performance advantages were demon- strated on the Apollo project,where crushable H/C tubecore and hexagonal cell H/C elements were used in the prim.ary and secondary struts, respec- tivl:::iy, of the Lunar Excursion Module landing gear.
Tubecore attenuator elements are formed by wrapping alternating layers of corrugated skin and face-sheet skin on a circular mandrel to obtain the desired ID/OD dimensions. The initial elements which were built and tested as part of the attenuator development program were fabricated in this manner from sheets of corrugated/flat skins, where the corrugated skin was brazed at each "wavelength" to the face sheet. The layers were not able to be brazed together during wrapping. This ".ll-:rnetal attenuator was design- ed to meet thf' requirement that the landing site not be contaminated with any organic compoL:.nds. At that point ~n time, the primary strut attenuator was composed of fi,re separate stages of tubecore, which were to crush at pre- scribed step-wise increasing force levels. The attenuator was assembled by bonding a spacer disc between each stage; the entire as sembly was about 40.6 cm long (four 1 in. stages and one high-force 12, in. stage). Crushing force of the stages increased progressively from 8230 nt. (1850 Ibs) to 21360 'nt. (4800 lbs).
These initial H/C attenuators, and several early variants, exhibited JPL Technical l-AenlOrandum 33-777 column instability problems during static and dynamic b~sting due to the ab- sence of brazing between wrapped layers. This behavior caused very erratic fluctuations in the crushing strength and unpredictable energy absorption capability.
On the basis of subsequent testing, it was concluded that a fulty-bonded tubecore attenuator design would perform properly. At about the same time, site alteration testing indicated .that descent engine shutoff at touchdown was ll feasible. Extensive "worst-case and statistical landing dynamics studies, similar to those conducted earlier, provided new H/C attenuator design re- quirements which optimized the trade-:.:>ffs between landing stabilh:y, shock;, and post-landed ground clearance. These new design requirements called for aUenuator s whose crushing load-vs- stroke characteristics would follow a ramped curve for the first 8.9 cm (3.5 in) of stroke, initially crushing at 4448 nt. (1000#), and then crushing at a constant 11200 nt. (2500#) force level for at least another 21. 6 cm (8.5 in.); these requirements are depicted in Figure 6. The tubecore vendor (Hexcel Corp. ) was able to supply attenuators whose static and dynamic crushing performance (also depicted in Figure 6) satisfied these design requirements. These attenuators were fabricated frolD fully bonded tubecore cylinders whose cross-section was tapered for 8.9 cm (3.5 inches) by machining; the elements were slightly pre-crushed, and then bonded to machined end-fittings for installat~('n in the primary strut.
The final flight-design attenuator assembly is about 53 cm (21 in. ) long.
These flight attenuators also have limited return- stroke capability to res ist tension loads up to about 445 nt. (100 Ibs). Figure 7 is a photo of flight-type attenuators; the upper assembly is shown compressed approximately 18 cm.
To conform with site non-contamination requirements, it was necessary to completely enclose these organically- bonded H/C aUenuator s r A laminate Kapton cloth contamination fairing, identified on Figure 2, was used for this purpose. An O-ring was used to seal off the enclosed attenuator as the prim- ary strut strokes; in addition, a microfilter was fitted to the top of the primary strut housing to provide venting during launch and landing. A long spring in the main strut tube is used to deploy the leg from its stowed configuration (Figure 8) after a pyrotechnlc pin-puller releases the leg about fifteen seconds following aero shell jettison. The main strut tube is treated with a Tufram coating to minimize deployment friction. A terminal engine shutdown switch (TESS) was fitted to each leg to provide shutdown commands to the lander computer. These switches are activated on touchdown as the leg cOlDpresses a weak spring a distance of 1. 3 cm (1/2 in. ) before any attenuator s tr 0 king ha soc c ur r ed.
222 JPL Technical Memorandum 33··777 CONCLUSIONS The final flight-design Viking landing gear geometry is shown in Figure 8. The evolution of this gear design has been described in terms of functional requirements defined primarily from analytical studies, and in terms of hardware performance (and problems) determined from Lest programs.
The more innovative features of this design are the footpad, the honeycomb main strut attenuators and the secondary strut load limiters. Their consistent performance, demonstrated in component development tests and in lander verification drop tests, perrn.itted landing dynamics analyses to be conducted with certainty.
The stringent and challenging ITlission requirements imposed on the landing gear have been satisfied by this design.
In particular: Probability of stability exceeds 99. 7~J ; Statistical ground cl~arance exceeds 22 cm; Three-sigma component landing shock is less than 30 g; The energy absorption capability of the gear elements permits landings at greater than three- sigma velocities; Site non- contamination and engine- shutdown requirements have been satisfied; and The footpad des ign permits landing on surfaces ranging from impenetrable to ver,! soft soils.
Furthermore, the total landing gear has evolved from an early design weighing over 45 kg (100 lbs) to the flight-des ign, which weighs only 20 kg (44 Ibs), less than 3.50/0 of the lander touchdown weight.
JPL Technical Memorandum 33-777 Figu re 1 - Original Gear Design I
st rut , /
~~ - Con tamination fai r in g
-
I
I
FiClure 2 - Final Fli ht G a r <:> JP L Techn i cal 1 morandum 33 - 777 , l' ; .
) " \ ! .
zr
4 , "'- .... .... 111"'",...
e 3 - Foo tpad Soil P nclration T sts F i gur 22 5 JPL Te chnic a l l lcmoranrlum 33-777
•
""1) ", 'si ,: i· 'oj" i ; I' i'illl'l III Figur e 4 - Load Limiter s STROKE - CM .~ ~ .- 2.0 .-' 7875 NT.
,-' .. - - V 1770 LB.)
--------::. .-
1-
-
V
~
-
-
"""""'-
- . -
1.5
==-====
~ - 60 .
-
I-
7 .- .-
~.~.~ Z
- ;.,
(-
r- REQU I REMENT C> C> ......
SCADER RANGE 8 ANi}-- 1.0 -
1/
§
Cl < o
I
-I .5 -
I
o 0
2 3 4 STROKE - IN.
Figure 5 - Load Limiter Requirement and Test Data .JPL Technical Memorandum 33-777 STROKE - C 10 15 3.0 5 r I I 1 177 0 NT (2 600 LB ) 12 0 DY AMIC 2.5 1001- Z eD ....J 2.0 8
§
0 0 « a ....J « r--- 1.5 a ....J REQU I REM NT
a
RA GE 1.0
- --
r-- I
I I I 1 I
d 6
2 3 4 5 0 I S TR O E - I T st Data nuator H >quirem nt and Figur 6 - Att 7 - Al nltalor \\' ilh E nd I· itting TPL Tt' chnic 1 1 'lllor lI1dum 33-777 22 7 N N (Xl -'I LEG :5 LANDER
-~O"
17.75 - \ PULLER OIA +~----r-~~'L-- -$J~-, +x
/1\0, ~_---;-;-;--;- __ --.t
(STOWED)
B-B
STROKE GAUGE
' '.. "-- CG /
- \
J
I \, STROKE GAUGE
I 1 -< --
J//; "0,:/'
BOTTOM OF ~-J,.. ,
, r CONTAMINATION _- LOAD LIMITER
\ FAIRING
CG 7
TOLR L' +--r-"-1"'~
z:t'5l--+-- -
- x , LEG 2 1\ +~
~ --
i 22 eM -1'----
IT 3.') ..- -7-, 'It!fIl'tmm '-FOOTPAD
- TEMPERATuRE : BIPOO-1
TRANSOllC EP rtt---
8-8 (LANDED)
---L-r----
....---,--
BOT rOM OF ----'- LANOtR BODY
B-B
F light Design Configuration Figure 8 -