Construction of a CubeSat using Additive Manufacturing

Construction of a CubeSat using Additive Manufacturing
Share on

Description of Satellite

The RAMPART satellite is a 2U CubeSat consisting of an upper BUS module, Lower Propulsion Module and Solar Panels.

Cubesat. CAD view of RAMPART with panels

CAD view of RAMPART with panels

The entire structure, built in Windform® XT via Additive Manufacturing, is plated in a High Phosphorus Electroless Nickel to provide radar reflectivity for tracking purposes. The upper BUS module consists of several experiments as well as test solar cell panels mounted to the exterior of the module. An additional experiment related to the performance measurement of the Windform® XT material has been placed at the top section of the BUS.

Top of BUS module showing load cell circled in red

Top of BUS module showing load cell circled in red

The load cell (Designed by Walter Holemans, Planetary Systems Corporation) measures the change in preload of Windform® XT. A 400 lb compression load in the load cell will cause 400 lb tension and about 2,500 PSI stress in the Windform® XT material. This integrated load cell is designed to measure creep or fracture as a result of exposure to radiation and thermal cycling over time.

  • The excitation will be 10.0V, response ~0.002 V
    Reponse x calibration factor = load in Windform® XT
  • The load will vary with temperature and time
    Lower temp = higher load

RAMPART’s Solar Panels will be spring–erected following the satellite’s deployment from its launch vehicle.

The aft portion of the satellite consists of a Micro Electrical Mechanical (MEMs) propulsion system designed and developed by Dr Adam Huang of University of Arkansas. The RAMPART propulsion system incorporates a miniaturized resistojet thruster core with a microfabricated de Laval nozzle and integrated heater. Upstream of the nozzle/heater assembly is an injector fed by 3 miniature solenoid valves. Prior to arriving at the valves, the propellant passes through a 2-phase separator membrane where only the gas phase of the liquid propellant can pass through its micron-sized pores. The membrane also serves as a filter for preventing the valve seats from becoming contaminated with debris. The propellant used is the refrigerant R-134a, which is considered nontoxic and nonflammable. The compressed fluid nature of the R-134a provides relatively high self-pressurization for delivery throughout the thruster system.

Although resistojets do not represent the state-of-the-art in thruster technology and offer relatively low thruster efficiencies (RAMPART’s design specific impulse of 90s), their simplicity and practicality apply well to pico- and nano-satellites. A key factor to the performance of the RAMPART propulsion system is the light weight and cell structure of the propellant tank made from Windform® XT. Instead of utilizing state-of-the-art technology, RAMPART uses high propellant mass fraction to provide the required Delta-V (320m/s for RAMPART) for Low-Earth-Orbit in-plane orbital maneuvers. Since the dimensions and the weight of a CubeSat are its main constraints, the Windform® XT allows these parameters to be optimized through the usage of multiple and interconnected near-cubic cells. This offers the advantage of maximizing the propellant volume, integrated baffles provided by the interconnecting walls, and possibility of correlating material strength tests of a unit cell to the entire propellant tank design. One can easily compare this design to traditional pressure vessels (hemispherical caps) and realize a two times difference in the propellant storage volume.

CubeSat. Top of propulsion module

Top of propulsion module

To determine material performance, test cubes similar in size to the baffled chambers, shown in Figure 3 and 11, were tested to various pressures. After physically testing the small test cubes that represented the chambers up to 600psi, an FEA model was developed by Whitney Reynolds of the U.S. Air Force Research Laboratory, Space Vehicles Directorate (AFRL/RV), to simulate the reactions of the small chambers and then correlate the results to a simulation of the final large propulsion unit.

Material Choice

The use of Rapid Prototyping Plastics in Aerospace has been somewhat limited to prototype work with the exception of Nylon 11 developed by Boeing (ODM) and Nylon 12 utilized by Northrop Grumman. (Programs include F-18 and Global Hawk). These two materials are used in a production environment and are produced using Laser Sintering (LS) machines. The LS technology has been seen as the best opportunity for durable plastic materials with a consistency in manufacture. Selective Laser Sintering resolution for plastics is commonly run at 0.004inch per layer utilizing a CO2 laser that can be adjusted to melt the plastics into a fully dense material.

Fused Deposition Modeling (FDM) by Stratasys, Multi Jet Modeling (MJM) by Objet, and Stereolithography (SLA) from 3D Systems are some of the Rapid Prototyping technologies examined. Each technology and material has benefits for prototyping. Key factors helped steer the material choice, including Heat Deflection temperature, UV exposure, and the requirement of plating the satellite in order to make it radar reflective.

Windform® was developed by CRP Technology based in Modena, Italy. Their interest in the materials was driven by a customer request for an improvement in stiffness – specifically for wind tunnel applications in Formula One racing.

F1 scale wind tunnel model equipped with Windform XT

F1 scale wind tunnel model equipped with Windform® XT

Windform® XT uses a base Polyamide and reinforces this with Carbon Microfibers. Racing teams were the first to begin using the material, and determined it to be a solution that could be utilized in “on car” applications. Windform® is currently used by both F1 (Figure 5) and NASCAR teams to replace components that would typically require injection molded materials.

Now Windform® XT has been replaced by the new Windform® XT 2.0 which is characterized by more performing features compared to XT. For this case-study was used Windform® XT as the new material XT 2.0 was under developing. The same cubesat is now manufactured with Windform® XT 2.0 with excellent result.

Windform XT used as brake inlet duct on F1 Car

Windform® XT used as brake inlet duct on F1 Car

Based on observations of other materials, samples of Windform® XT were subject to Tensile Test and the cross sections of the brakes were examined under an electron microscope. The micrographs showed the carbon microfibers were encapsulated by the Nylon base material. In addition little to no porosity was visible in the internal structure.

Courtesy Paramount Industries

Courtesy Paramount Industries

In addition, studies have shown that Windform® XT performs in a predictable manner and has been compared to injection molded production materials to determine how it would react in exposure to extended temperature cycling. Having a predictable material is important in the examination of possible failure modes for design studies.

Comparison of Windform XT to PA6 BG-35 injection molded plastic

Comparison of Windform® XT to PA6 BG-35 injection molded plastic

The chart above reporting the evolution of tensile strength between the comparison of Windform® XT and PA6 BG-35, makes Windform® XT a good candidate for Additive Manufacturing for the CubeSat application.

  • Windform® XT passes Outgas Screening, ASTM E-595
  • Windform® XT is produced in a manner that makes it dense
  • Windform® XT can be easily machined using conventional methods
  • Windform® XT has a good heat deflection temperature (HDT) relative to other RP materials. (HDT above 170C)
  • The base polyamide material has been proven to meet performance needs for other Aerospace applications
  • Material batches are quality controlled, each coming with a Certificate of Conformance (COC)
  • Build volume of SLS system fits well with CubeSat applications= 381 x 330 x 457 mm (14.5 x 12.5 x 17.5 in)
  • Windform® XT can be plated without the need for sealing agents

Construction of BUS and Propulsion Module

Windform® XT is used to build parts using the Laser Sintering process. This process can be described in several steps.

Illustration of LS process

Illustration of LS process

Laser Sintering is a powder based Additive Manufacturing method. The process uses a layer system of building up a part with each layer of powder being sintered separately by a laser. As each layer is sintered it slowly constructs the part step by step. There are various grades of materials used each with their own characteristics, but each are used the same way within the LS machine.

The LS machine basically consists of three powder beds and a Laser. Two of the powder beds hold the feed powder and the third bed holds the part. The part bed is in the middle of the beds with the laser acting directly perpendicular to this bed. A roller is used to push the layers of powder over the part bed and all three beds have their own heater source. The process itself is a very simple, repeatable one.
The building of the parts is a repeatable three-step process.

Step 1

A roller is positioned beside one of the feed beds. This feed bed then raises a set amount (Usually < 0.1mm) and the roller pushes the raised powder across, covering the part bed with a Powder Layer.

Step 2

With the layer of powder present, the laser starts to trace out the desired shape of the part in the powder, melting the powder as it contacts the surface.

Step 3

Once this is done, the part bed drops down the set amount and the process continues from the opposite side, with the other feed bed raising and the roller distributing another layer of powder over the part bed, and the laser again traces the shape.

Steps of LS process

Steps of LS process

The part is built up in slices, with each layer of powder representing a single slice of the part. As the laser melts the powder, each layer fuses together to grow a solid part. Because of the fact that the part is made up in layers, very complex shapes and design can be manufactured that would otherwise be impossible by conventional means.

LS Process benefits

  • Layered technology is free from geometrical constraints: it is possible to build undercuts, hollow parts and internal ducts.
  • Opportunity to build directly functional “assembled mechanisms” (minimum clearance between parts 0.5 mm).
  • It is possible to build many different parts together in the same building volume.
  • Building time is not dependant on object geometry, but only on part volume and on build “height” (Z dimension).
  • Building time is very short (max 1-2 working days).
  • Lead time is very short ( max 2-3 working days).
  • When a great dimensional accuracy is required, Windform parts can be machined.
  • To reduce mass and weight it is possible to create hollow parts with internal reinforcing structures.

Items to consider when designing for SLS Process

  • Building volume (360x310x380 mm). In order to build big objects, it is necessary to produce piece by piece and then bonded together with approved materials.
  • Minimum feasible wall thickness is 1 mm. Ideal value is 1.5-2 mm.
  • Minimum feasible details >=1 mm.
  • It is important to avoid enclosed volumes: hollow parts will need a hole or some type of gate to allow the removal of loose or un-melted powder.
  • When designing parts to be assembled together always keep a minimum 0.2 mm clearance between them.
  • Excessive wall thickness (>10 mm) can cause unwanted warping and shrinkage on parts.
  • Roughness after LS process is high (Ra 6 m) and Z layering is visible. It is possible to smooth external surfaces to reduce it to Ra 1.5 m.

Windform® XT uses micro fibers as a reinforcing system and is similar to carbon lay-up techniques in that it is non-Isotropic. Similar design rules are applied when building parts to take advantage of the increased strength in one direction over the other resulting from the build process.  In the case of the RAMPART BUS, the decision was made to produce the part with the Z build axis vertical to the satellite’s longest axis after assembly.  This allows the X and Y axes to be positioned to provide the most strength in the perpendicular axis holding the BUS together.

BUS Construction

The BUS underwent several revisions within the period of a few months.  Walter Holemans was able to add experiments, create different deployment options, and develop wire routing and harness control devices utilizing design techniques to maximize the use of Additive Manufacturing.  Increased complication was not a hindrance to the process and allowed for greater design freedom.  The Solar Cell experiment (Figure 10) was easily tucked into the side of the BUS by only needing to implement a simple feature cut in the CAD system.  A new module design was built with the space in the side of the module and allowed the experiment to be added quickly to the mission.

Solar Cell experiment added to RAMPARTs BUS module

Solar Cell experiment added to RAMPART’s BUS module

The Bus and Propulsion Modules were created in separate sections to allow the different teams to test and assemble them at different locations.  This allowed the production of prototype modules to try ideas based on the current challenges and then adjust as needed.  At the completion of the final design, the parts were grown at 0.004inch (0.1mm) layers and then sanded to a smooth finish to aid in the plating process.

Utilizing the experimental and simulation data, the Propulsion module (Pressure Vessel) was oriented such that the Z build orientation was again parallel to the longest axis of the CubeSat.  Due to the internal structures of the baffled cubes, the module was subjected to an ultrasonic water bath to ensure no material was trapped in the chambers.
The plating was completed by Quaker City Plating located in Whittier, California, under the direction of Frank Huizar.


The innovative use of Additive Manufacturing has shown that this type of technique will allow for fast adaptability and freedom of design. Working with the idea of utilizing additive manufacturing as a core principle, the team was able to modify, change, and add experiments without concern for having to develop tooling or modify an existing cube structure.

The development of RAMPART  shows that this type of building technique adds value for standardized CubeSat boards as well as for devices that do not fit easily “in to the box”. As electronics and sensor systems become smaller and more complex, there is a push for developing more complex and advanced CubeSats. In the development of greater complexity, Additive Manufacturing opens the possibility to adapt the structure to carry new sensors, or optics. Utilizing CAD and additive manufacturing, the internal structure of the CubeSat can be built to adapt to the components, instead of the other way around. This allows the use of the CAD system to adapt to new technology as it is introduced. This eliminates the concern for legacy systems that are being held on the shelf, for technology that is becoming obsolete.

In addition, standard components can be placed into a CAD library that will allow for parametric generation of portions of the satellite. Additive Manufacturing of the RAMPART BUS module shows a clear mixing of standard board modules and customizing for solar panels , wire routing, and the addition of the load cell. Another observed benefit was the reduction in fasteners and ease of assembling the RAMPART CubeSat.  Positioning features and “snap fit” devices can be incorporated into the design methodology to speed assembly.  The extensive baffle design in the tank structure of RAMPART leverages this consolidation of components, removing the need for a complicated assembly.


Franco Cevolini
CEO CRP Technology S.r.l.
Walter Holemans
Planetary Systems Corporation
Adam Huang
Department of Mechanical Engineering, The University of Arkansas
Stewart Davis
Director of Operations CRP USA, LLC
Gilbert Moore
Project Starshine