American Institute of Aeronautics and Astronautics

1

A Multi-payload Adapter for Peacekeeper-based Space Launch Vehicles

2Lt Bryan J. Fram1

United States Air Force Research Lab., Kirtland AFB, NM, 87117

and

Gareth R. Thomas2 and Cynthia M. Fadick3 ATA Engineering Inc., San Diego, Ca, 92130

A key enabler of many future DoD Space Test Program (STP) missions is a low cost, reliable launch vehicle with the capability to place multiple payloads, of moderate mass, into orbit. The Peacekeeper Space Launch Vehicle (PKSLV), being developed by the Rocket Systems Launch Program (RSLP) of the Space and Missile Systems Center (SMC), utilizes decommissioned Peacekeeper ICBM assets to economically place payloads into orbit. Depending on configuration and launch site, the PKSLV can place up to 1590 Kg into Low Earth Orbit (LEO). Growth versions of the PKSLV may be able to launch up to 3500 Kg. To get maximum usefulness and economy out of this launch capability, multiple payloads will often need to be launched on a PKSLV. One potential drawback to the PKSLV is the severity of the launch environment experienced by the payload(s). Therefore, any sensitive payloads that fly on the PKSLV will likely require some form of vibration isolation and shock damping system. Under a Phase II Small Business Innovative Research contract, with the Air Force Research Laboratory – Space Vehicles Directorate, ATA Engineering is developing a Multi-Payload Adapter (MPA) for the PKSLV. ATA’s MPA, on current versions of the PKSLV, will be able to support a typical manifest comprising of a primary payload of approximately 1,000 lbs and four secondary payloads of up to 200 lbs. The MPA design consists of an annular flat plate that has top and bottom face sheets separated by radial ribs and close-out rings. These components are manufactured from graphite epoxy composites to ensure a high stiffness to weight ratio. The design is tuned to keep the frequency of the axial mode of vibration of the payload on the flexibility of the adapter to a low value. This is the main strategy adopted for isolating the payload from damaging vibrations in the intermediate to higher frequency range. The design challenge for this type of adapter is to keep the pitch frequency of the payload above a critical value in order to avoid dynamic interactions with the launch vehicle control system. This high frequency requirement conflicts with the low axial mode frequency requirement and this problem is overcome by innovative tuning of the directional stiffnesses of the composite parts. A second design strategy that is utilized to achieve good isolation characteristics is the use of constrained layer damping. This feature is effective at keeping the responses to a minimum, in particular for the resonant burn condition present in any stage powered by a solid rocket motor. ATA is currently entering the detailed design phase of the MPA development. By early 2005 ATA will have prototype flight hardware manufactured for the PKSLV MPA. At that time, qualification testing to confirm the MPA’s structural performance will commence. A potential first flight for the MPA is in the 2006-2007 time-frame.

I. Introduction s a new generation of space launch vehicles becomes operational, new challenges arise to safely, and cost

effectively, put payloads into orbit aboard them. Decommissioned Peacekeeper ICBMs, refurbished into the Peacekeeper Space Launch Vehicle, present some interesting technical challenges. The most important challenge is to design a payload adapter that can protect a payload from the harsh vibration environment experienced during launch. Also, the lift capability, in mass and volume, of Peacekeeper-based systems naturally leads to the desire to launch multiple payloads. Maximum lift capability on a current PKSLV is 1590Kg to LEO with a payload envelope of 80.9’’ diameter by 121.6” long. Growth potential exists for the PKSLV to launch up to 3500 Kg. Designing and fielding a multiple payload adapter with built in vibration damping to take advantage of this capacity is the goal of collaboration between the DoD Space Test Program, the Rocket Systems Launch Program, the Air Force Research Laboratory, and ATA Engineering Inc. STP, chartered to launch as many DoD space experiments as possible, can 1 Tech Lead, Advanced Spacecraft Mechanisms, 3550 Aberdeen Ave SE, Kirtland AFB, NM 87117, Member. 2 Vice President, Design/Analysis, 11995 El Camino Real, San Diego, CA 92130. 3 Lead Engineer, Design/Analysis, 11995 El Camino Real, San Diego, CA 92130.

A

Space 2004 Conference and Exhibit28 - 30 September 2004, San Diego, California

AIAA 2004-6124

This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States.

American Institute of Aeronautics and Astronautics

2

use this new adapter to place multiple experiments into orbit with one launch. Commonality between the interface for secondary payloads on the MPA and the EELV secondary payload adapter, ESPA, can lead to significant cost savings, and the potential for plug-and-play, due to payloads needing to design only to the standard interface, not a particular launch vehicle. The cost of a Peacekeeper-based launch, at approximately $20M, is much more economical when split amongst multiple payloads. ATA’s design will guarantee all payloads a softer and safer ride to orbit.

II. Technical Background Payloads, such as satellites or spacecraft, which are mounted on launch vehicles, are subject to severe vibrations

during launch. These vibrations are induced by multiple sources from launch to final separation. The dynamic mechanisms include ignition and operation of the rocket engines, transient vectoring forces at the nozzles, separation of rocket stages, aerodynamic effects and acoustic phenomena. The vibrations are often associated with severe quasi-static loads caused by axial thrust. The frequency content of the vibrations generally extends from 10-20Hz to several kHz. The amplitude of the vibrations tends to be more severe in certain frequency bands and this is usually a function of the type of rocket motor being used. The PKSLV solid rocket motors generate high vibrations in the 55-70Hz range.

A direct result of the severe vibrations generally experienced by payloads is that fatigue damage and failure can be incurred by sensitive payload components. Studies done by NASA in the 1970s1,2 showed that of all first day satellite failures, 30-60% could be attributed to launch vibration. Extensive engineering effort is normally expended to ensure that this phenomenon is fully understood and avoided. Up to 70% of payload structural mass is dedicated solely to ensure that the payload survives the launch loads.

The mounting of the payload to the launch vehicle is usually done by attaching the lower spacecraft interface to the forward end of the rocket. Thus the payload is normally cantilevered at the front of the launch vehicle. This configuration leads to axial components of the interface forces between the payload and vehicle even in the presence of purely lateral loads. These components are additive to those caused by the axial loads and vibrations. This observation underlines the importance of the axial load transfer at the interface.

A rigid connection at the payload/vehicle interface has been widely used in the past especially for vehicles with very robust payloads. In situations where a few sensitive components are to be used in the payload, these components are sometimes attached using vibration isolation mounts. This approach is not cost or weight efficient for a fragile payload and the concept of complete payload isolation is now widely adopted in such situations.

Complete payload vibration isolation schemes generally use a flexible payload/vehicle interface. When the natural frequency of the payload vibrating on the flexibility of the interface is significantly lower than the frequency of the vibrations being transmitted through the vehicle to the interface, the payload is essentially isolated. The real challenge in the design of a satisfactory complete payload vibration isolation system is to satisfy two competing requirements. First, a payload mounting frequency low enough to achieve good isolation is required. Second, it is important to avoid the problematic interaction of the low frequency payload modes of vibration with the low frequency primary bending modes of the vehicle. The first requirement drives the payload frequency down while the second drives it up.

The difficulty of satisfying the two competing requirements is best understood with reference to specific frequencies. In cases where it is desirable to isolate 55Hz vibrations, the axial payload mode must be less than 39Hz to get any attenuation at all. A frequency of about 25Hz would be desirable as this would achieve a vibration transmissibility of only approximately 27%. The 25Hz value is a lower bound of the acceptable frequency range because of interaction problems with rocket axial modes of vibration for a broad range of launch vehicle designs. Therefore, for these rockets, the payload isolation frequency in the axial mode of vibration should be in the 25-39Hz range.

The lateral modes of vibration of the payload make the frequency requirements even more difficult to satisfy. Problematic interaction with bending modes of the rocket leads to a common requirement that the payload lateral mode of vibration should be greater than approximately 15Hz. The lateral and axial modes of vibration of the payload are generally closely coupled for typical adapter designs. It is noted that if discrete translational springs are used to introduce flexibility into the payload mounting system, it is difficult to avoid having the pitch mode lower than one third the bounce frequency. This ratio of one third is applicable for geometries where the mounting circle has a diameter of approximately 62in and the payload center of gravity is 60in above this circle. Thus the 15Hz pitch mode may well be associated with a bounce frequency of 45Hz which is clearly too high to achieve attenuation of the 60Hz vibrations. These vibrations may well be amplified for such a system rendering the concept of vibration isolation infeasible for the frequencies cited. Only if frequency ratios of considerably more than one third are

American Institute of Aeronautics and Astronautics

3

achieved does the concept of vibration isolation become possible. A graphical interpretation of the competing frequency requirements is shown in Figure 1.

Figure 1. Payload adapter design space as a function of pitch and bounce frequencies showing two families of

designs with frequency ratios of 0.29 and 0.5.

Different payload adapter designs can be represented as points in Figure 1 and a family of designs based on a particular concept would be represented as a line. Two such families are shown with lines that have constant ratios of pitch frequency to bounce frequency of 0.29 and 0.5. Areas shaded in red in the figure represent designs that are infeasible for a variety of reasons:

• Bounce frequency too low (below 25Hz shown) • Pitch frequency too low (below 12.5Hz shown) • Bounce frequency too high leading to inadequate vibration isolation (above 35Hz shown)

This plot emphasizes the importance of frequency in the design of a payload adapter for the PKSLV. Another parameter that has significant bearing on the performance of a vibration isolation system is damping.

Typically the higher the damping the greater is the vibration attenuation. It is therefore desirable to incorporate damping features into the design of an isolation system as is done in the present development.

III. Payload Adapter Design The thrust of the present development is to use a concept for payload mounting that avoids the high axial-to-

lateral frequency ratio inherent in the commonly used discrete spring isolation concepts. A further goal of the present development is to use hardware that provides a convenient and integral mounting scheme without resorting to add-on devices such as springs or flexures. The proposed strategy leads to cost and weight savings.

The present payload adapter invention consists essentially of a flat annular plate or disk (Figure 2) that has inner and outer bolt circles. The outer circle facilitates a flanged connection to the launch vehicle upper stage while a payload is attached at the inner circle. When multiple payloads are accommodated, an adapter is mounted to the

American Institute of Aeronautics and Astronautics

4

plate at the inner bolt circle. The special isolation characteristics of the disk, effective for all payloads, are achieved by means of specific construction features.

Figure 2. Payload adapter design concept with no payload shown

The annular plate consists of two face sheets separated by radial ribs and inner and outer closure rings (Figure 3). The use of isotropic materials for the various skins would lead to a payload adapter that has a ratio of axial frequency to lateral (pitch) frequency of more than three. In other words, this device would not be effective at isolating vibrations for excitation frequencies of 50-60Hz unless the pitch mode was at a prohibitively low frequency. The desired effect of moving the axial and pitch frequencies closer together is achieved by the use of highly directional material properties. This directionality is an integral characteristic of the proposed composite materials and the use of such materials is critical to the design.

Accommodates upper stage bolt circle

Payload cg Lightening hole

Accommodates payload bolt circle

American Institute of Aeronautics and Astronautics

5

Figure 3. Payload adapter with top skin removed to expose radial ribs

The relative change in stiffness that is sought in the design is the softening of the axial (bounce) stiffness and the stiffening of the pitch stiffness relative to an isotropic material version of the design. This is done by the use of highly directional composite fiber laminates. The use of composite face-sheets where the fibers are concentrated in directions of +/-45o relative to radial lines will lead to very low material stiffness in the radial and hoop directions. The principal strain directions for a pure axial mode of the payload adapter are shown in Figure 4 and are either radial or tangential. Fibers oriented at +/-45o will add virtually no contribution to the stiffnesses in these principal stress directions. Consequently the resulting stiffness in the axial direction will be small. Of course, this effect is beneficial only if the pitch stiffness does not reduce correspondingly. In other words we are seeking a method by which the axial stiffness is reduced more than the reduction in pitch stiffness.

American Institute of Aeronautics and Astronautics

6

Figure 4. Principal stresses in top skin (shown with colored arrows) are either radial or tangential for the

axial mode of vibration

The +/-45o fiber configuration produces significant stiffness for the pitch mode of deformation. This can be best understood by reviewing the principal stress directions for the face-sheets for that mode of deformation (Figure 5). It can be seen that the principal strains are at +/-45o along the radial lines that remain straight during plate bending in this mode. These lines are called nodal lines. In other words, the +/-45o fibers located close to the nodal lines provide significant stiffness in this mode.

Deformation of top skin in axial mode of vibration

Principal Stresses

American Institute of Aeronautics and Astronautics

7

Figure 5. Principal stress direction in the top skin is at +/-45o to the radial nodal lines for the pitch mode of

vibration

This discussion has shown that the use of a composite material with fibers oriented at +/-45o to radial lines produces very little stiffness in the axial mode of vibration while providing some stiffness in the pitch mode. The realization of the +/-45o orientation consistently for all radial lines requires special attention and details of the proposed approach are beyond the scope of this paper.

Radial stiffeners are apparent in Figure 3 and their role in the device is explained with reference to the axial and pitch modes of deformation given in Figures 4 and 5. It is again noted that the use of isotropic properties for these ribs would lead to significant stiffness in both modes of vibration. This is not a desirable outcome since decreasing bounce stiffness more than pitch stiffness is the objective of the design.

The radial ribs will also be made from composite materials with fibers oriented at +/-45o to the long rib edges. Ribs configured this way will have virtually no stiffness for the bounce mode of vibration. The top and bottom edges of the ribs will be able to extend or compress without straining the inclined fibers in this mode of deformation.

The ribs in close proximity to the nodal lines in the pitch modes of vibration (Figure 5) will be subject to vertical shear deformation for which the +/-45o fibers will contribute significant stiffness. Another way to interpret this is to imagine the top and bottom skins and radial stiffeners making a series of imaginary radial box beams connecting the inner and outer adapter flanges. These beams will have very little bending stiffness but will be quite stiff in torsion.

Nodal line

Top skin deformation in pitch mode

Nodal line along which there is no

normal displacement

Principal stress arrows

Direction of payload motion in pitching mode

American Institute of Aeronautics and Astronautics

8

Bending is the only mode present in the beams in the bounce mode of vibration. The pitch mode will put some of the radial box beams in torsion, hence providing some stiffness while the beams at right angles to the nodal lines will be in bending and provide no stiffness. However, the net result is that pitch stiffness is achieved without a corresponding increase in bounce stiffness.

The beneficial effect of damping has already been noted and this effect is utilized in the present device. The radial ribs seen in Figure 3 are in fact pairs of ribs located in close proximity. One of the ribs in each pair is attached to the top skin while the other rib is attached to the bottom skin. The use of a constrained layer damping material between the two ribs is an ideal location for enhancing the damping in both bounce and pitching modes of vibration. There is a tendency for the adjacent ribs to shear relative to each other in either mode. There is no corresponding tendency to separate the two ribs in these same modes. Thus the damping material, despite having weak tensile strength, is unlikely to fail in tension.

The discussion so far has addressed a single payload configuration. This can be easily adapted to accommodate multiple payloads using the secondary adapter shown in Figure 6. For the single payload configuration only the payload adapter disk is required whereas for multiple payloads a secondary adapter is needed. This design achieves isolation for all payloads from the PKSLV launch loads. In addition, the modularity allows an off-the-shelf payload adapter disk to be used in combination with a custom-designed secondary adapter. This approach limits the customization to the relatively straightforward secondary payload support while taking full advantage of the structural sophistication designed into the payload adapter disk.

Figure 6. Modular payload adapter design accommodates single and multiple satellite manifests

Primary payload interface

Single Large Primary Payload

Secondary payload

interfaces

Multiple Payload Manifest

PK upper stage

PPaayyllooaadd aaddaapptteerr

ddiisskk

Primary payload

Secondary adapter

PK upper stage

Secondary adapter

Secondary payloads

Primary payload

PPaayyllooaadd aaddaapptteerr

ddiisskk

Adapter cone

American Institute of Aeronautics and Astronautics

9

In summary, the design concept that is described here has the capability of allowing independent tuning of the pitch and bounce modes of vibration for single or multiple payloads. Certain material and geometric parameters can be selected to move the modes to desirable natural frequency combinations. This can lead to superior complete payload isolation without compromising the rocket dynamic performance vis-à-vis stability and control issues. The characteristics of the adapter are achieved without deviating from relatively standard composite material and construction technology. Furthermore, the concept can lead to a durable, strong and lightweight hardware.

IV. Analysis The feasibility of the proposed design concept has been verified by a series of finite element analyses. The model

used for these analyses, as shown in Figure 7, uses a rigid representation of the payload which is adequate for preliminary studies.

Figure 7. Finite element model of payload adapter and the stresses induced for a static load case (10g –

axial, 5g – lateral for a 2,500lb payload) Figure 7 also shows the stresses induced by a critical static load condition used to qualify the strength of the

structure. This case corresponds with inertia loads of 10gs in the axial direction and 5gs lateral. Also of interest as discussed earlier is the modal behavior of the adapter; this is shown in Figure 8.

Figure 8. The first two modes of vibration of the adapter and payload

The modal response of the system falls within the general guidelines discussed earlier. Of particular note is that

the ratio of the pitch frequency to the bounce frequency (0.54) is significantly greater than the one-third value cited previously. This high value is testament to the effectiveness of the stiffness tuning discussed in the Section III.

The final set of results to illustrate the adapter’s vibration isolation capability is shown in Figures 9 and 10.

First bending mode 14.5 Hz

First bounce mode26.8 Hz

American Institute of Aeronautics and Astronautics

10

Figure 9. The axial acceleration response at the payload cg for the random input indicated (in green) for

three different payload masses

The straight line data in the plot of Figure 9 is the base excitation which has a 3σ value of 38.3gs. The 3σ value for the payload cg is only 1.9gs even for the lightest payload considered, which incidentally is the most difficult to isolate. Similar response results have been achieved for the multiple payload manifest scenario. Figure 10 shows a comparison between the acceleration transient responses at the CGs of the primary and secondary payloads. These results were generated in a transient dynamic analysis in which the launch vehicle interface was excited with vibrations consistent with the PSD shown in green in Figure 9.

Figure 10. Transient axial acceleration response at the cgs of primary and secondary payloads. The red

lines represent the input acceleration.

2500 lb payload 3σ response = 1.38G

3σ base input = 38.3G

1600 lb payload 3σ response = 1.90G

2800 lb payload 3σ response = 1.27G

Primary Payload Secondary Payloads

American Institute of Aeronautics and Astronautics

11

V. Conclusion This paper presents a design concept for a payload adapter that can be used for single or multiple payload

configurations on a PKSLV. The emphasis of the development has been on the vibration isolation performance of the adapter and its ability to support multiple payloads. Isolation performance is accomplished by keeping the bounce frequency of the payload on the flexibility of the adapter as low as possible while keeping the frequency above the constraints imposed by vehicle system requirements. Two conditions need to be addressed that conflict with the bounce frequency requirement. First, the static strength due to axial loads needs to be maintained at a high level despite making the corresponding stiffness low. Second, a high payload pitch frequency must be maintained while simultaneously minimizing the bounce frequency. These conflicting requirements are satisfied in the present development through the innovative use of directional properties in a composite lay-up. The results of extensive static and dynamic analyses are reported to demonstrate the effectiveness of the design. The next major step in the overall development program is to build and test prototype hardware to confirm the feasibility of the design. 1 Timmins, A R. “A Study of First-Month Space Malfunctions.” NASA Technical Note D-7750, 1974. 2 Timmins, A. R. and Heuser, R E. “A Study of First-Day Space Malfunctions,” NASA Technical Note D-6474, 1971.