LV3: Design and Manufacture of an Open-Hardware Rocket

Airframe using Carbon FiberProgress Report

Mar. 2016

Advisor:Dr. Mark Weislogel

Members:Leslie Elwood

Erik Nelson

Joseph Shields

Jacob East

Brandon Bonner

Table of Contents

Introduction 2

Mission Statement 2

Project Plan/PDS 3

Customer Identification 5

Conclusion 6

Appendix

Carbon Fiber Layup Procedure 8

Nosecone Selection 15

1

Introduction:The Portland State Aerospace Society (PSAS) is an interdisciplinary group of engineering

students and alumni of Portland State University (PSU) with the long term goal of putting a

cubesat into orbit with their own rocket.

Their current airframe, named Launch Vehicle 2 (LV2), has served for over 12 years,

representing 10 of the group's 13 launches, and hosted experiments ranging from custom patch

antennas and long range WiFi technology to GPS navigation and a cold gas reaction control

system. The LV2 platform is mostly constructed of aluminum with a fiberglass shell, with many

of the parts having been fabricated in home garages. This makes for a robust but heavy design.

Additionally, this airframe is built with a 4.5 inch inner diameter which PSAS's experiments have

outgrown.

The new airframe being designed, named Launch Vehicle 3 (LV3), aims to address these

issues. The LV3 platform uses a 6 inch inner diameter, modules composed of carbon fiber and

thin aluminum coupling rings, a carbon fiber nose cone, and a carbon fiber fin section. All of the

airframe components connect via standardized rings, to accommodate future experimental

modules and flight configurations.

The cylindrical LV3 airframe modules already outperform the old design with an 80% reduction

in weight.

Mission Statement: The primary goal of the LV3 project is to reproduce the process the 2014 capstone team

developed for manufacturing the cylindrical carbon fiber sections of the rocket to be used on

future PSAS projects, and to design and manufacture a nose cone and fin section using the

same technology. We will clearly document our design and manufacturing processes, so that

future PSAS can easily produce cylindrical modules, nose cones, and fin cans as needed. A

summary of the nose cone selection and the manufacturing process that we have implemented

to date is detailed in the appendix.

Project Status:

2

The status of all the project goals can bee seen in Table 1, below. In short, the LV3 team has acquired nearly all of the necessary material and machine time donations, is receiving advice from a composites professional at Boeing, and has completed three attempts at reproducing the previous team’s results.

Table 1: status of project goals.

Additionally, the time-dependent status of the project is as follows. Starred items have been

delayed past their originally planned date.

Task Date

3

Carbon Fiber Practice layup 1/27 (done)Material selection

Airframe 1/29 (done)Cone 1/29 (done)Tailcan 1/29 (done)

DesignNosecone 3/25 (done)Fins 2/12 (done)Nose Cap 4/12*

Material AcquisitionAluminum 1/29 (done)Carbon Fiber 2/19 (done)Nomex 2/19 (done)Scotch-weld Adhesive 2 2/19

(done)Manufacturing

Couplers 2/19 (done)Airframe 4/1*Tailcan

Fins 4/10*Base 4/1*

Nosecone 5/18*Testing

module crushing 3/25fin bending 4/15nose heating 5/25

MiscellaneousMoving data from L12 to L13 repository 1/29 (done)L13a launch 7/15L13b launch 7/16

The material acquisition and machine time donation could have potentially left the LV3 project

dead in the water. With these tasks complete, we have moved on to the design and

manufacturing phases of the project. Already, three attempts have been made to reproduce the

results of the previous capstone team.

Presently, we are finalizing the design of the nose cone to prevent overheating of the epoxy in

the carbon fiber (see appendix). The aluminum frames which complete the aerodynamic profile

of the fins are being produced by the Machine Sciences Corporation as a donation to the group.

We are pursuing further donations of adhesive film, which is needed to complete all the parts

4

required by PSAS. Additionally, we are receiving advising from professional composites

engineers at Boeing.

Due to recent changes in the expected launch schedule, the LV3 airframe must be ready for two

different launch configurations in the summer. The first is the stress test launch (L13a), which

will test the stability and durability of the LV3 airframe. It will fly with no payload, and the largest

motor available. The second launch will fly with a small payload and include the roll control

module. This scenario is a much closer approximation to the future flights of the LV3 airframe.

Customer Identification:

External customers

We will be delivering our products and documentation to Andrew Greenberg and the Portland

State Aerospace Society, in fulfilment of our capstone project.

Andrew is the general manager of PSAS, a successful entrepreneur, leader of Portland State’s

business incubator, and an adjunct faculty member at PSU. He oversees the general success of

PSAS and coordinates their many interdisciplinary projects. The LV3 team is working closely

with Andrew to ensure our products and documentation meet the specific requirements of the

many other projects flying on the LV3 airframe.

The Portland State Aerospace Society is the interdisciplinary rocketry club at PSU. PSAS plans

to eventually put a small cubesat into orbit. In pursuit of this goal, they must move to smaller

and lighter airframes which they can later modify and replace. Not only do they require a

functioning airframe from the LV3 team, but also clear and detailed documentation in order to

adapt the LV3 designs to future projects.

Internal customers

The internal stakeholders of the LV3 project are the team members and their advisor, Dr. Mark

Weislogel.

5

All of the LV3 team members are senior mechanical engineering students who will be moving

into aerospace related firms or graduate studies. Each member is invested in the success of the

LV3 project, and we look forward to the future launches of the LV3 airframe and impacts of our

designs.

Dr. Mark Weislogel, a professor at PSU and former NASA engineer, is serving as the advisor for

the LV3 project. He evaluates the progress of the team and documentation they produce.

Conclusion: The LV3 team has come a long way since the initial product specification report. Material

acquisition, which was the biggest concern at the start of the project, has been completed after

weeks of identifying and soliciting possible donors for the material we needed. During that time

we educated ourselves on the manufacturing processes and the parameters affecting our

design requirements. We recieved coupling rings and a promise of continued support from

Machine Science Corp. in Wilsonville. We also shared this resource with the LV4 and the liquid

fuel motor teams. Pacific Coast Composites, based in Lakewood, Washington, donated

approximately 15 thousand dollars worth of aerospace grade pre-impregnated carbon fiber,

fiberglass, and 3M adhesive film. Since then, we have completed two module layups that are

currently taking approximately 14 hours from mandrel preparation to removal of the module from

the mandrel.

A major issue that will need to be resolved in the coming weeks is the lack of 3M adhesive film.

Seventeen Yards was donated from Pacific Coast Composites, but this is not enough to

complete our project goals. We expect that there will be wasted material due to our gaining

experience with this manufacturing process. More 3M adhesive film will either need to be

obtained via donation or purchased.

In order to begin the manufacturing of the nose cone and fins it is essential that current surface

finish issues are resolved, so material isn’t wasted on bad layups. We are currently working with

the established cylindrical module process to correct for differences in the material we were

donated versus what was used to create the original process.

We have met our goals thus far and later due dates are reasonable. The materials have been

procured, industry relationships have been established, and other material not donated to us

6

has been located and priced. The project is on track to finish on time, assuming more adhesive

film can be obtained.

7

AppendixCarbon Fiber Layup Procedure

The LV3 team completed three layups of cylindrical modules using carbon fiber. The first

attempt could not be removed from the mandrel without being destroyed. We gained practice

with the layup process, which made the next layup faster. The second layup was able to be

removed from the mandrel by hand. It has wrinkles in it that we are not sure how they were

created. Below is a detailed process of our second carbon fiber layup.

Materials:● orca skin ● orca seal ● 320 grit sandpaper● 600 grit sandpaper● 1500 grit sand paper● acetone● powderless nitrile gloves● insulated gloves (gardening gloves will work)● a small piece of acrylic (optional)● popsicle sticks or tongue depressors● paper towels● Utility knife● Scissors

Mandrel preparation:In figure 1 below is shown the waxed mandrel along with a container of Orca Skin, and Orca Wax.

● use a piece of acrylic or a popsicle stick to scrape away any chunks of epoxy or adhesive from the previous layup

● wet sand with 320 grit paper, to remove the remaining epoxy and adhesive from the previous layup

● wet sand with 600 grit paper, removing the marks from the previous sanding● wet sand with 1500 grit paper, removing the marks from the previous sanding (figure 1)● put on powderless nitrile gloves● wipe with acetone and a paper towel until the towel comes back clean● apply 2 coats of orca seal● apply 7 coats of orca skin, 10 minutes between coats● wait 1 hour

8

figure 1: the waxed mandrel on the preparation table, along with a container of Orca Skin, and Orca Seal.

Dummy ring preparation:● wet sand with 600 grit paper to remove any stray spray adhesive from the previous layup● clean with acetone or water

Coupling ring preparation:● place the mandrel on the mandrel-holder● put on powderless nitrile gloves● wet sand the adhering surfaces with 600 grit paper ● clean all surfaces with acetone ● continuously apply the anti-galvanic 3M-130 part A to the adhering surfaces for 3

minutes● continuously apply the anti-galvanic 3M-130 part D to the adhering surfaces for 3

minutes● take care not to touch the adhering surfaces from this point on● apply 1 coat of orca seal to the innermost surface● apply 3 coats of orca skin to the innermost surface, waiting 10 minutes between coats● wait 1 hour

Cutting films and fabrics:● select the appropriate cutting template for the desired film or fabric● lay the film or fabric over the plastic surface of the work bench● place the template over the film or fabric● while one person moderately presses down on the template, another person should cut

along the edges with a utility knife

9

When cutting the overexpanded nomex honeycomb core material, make sure the template is oriented correctly. You should only be cutting single rows of cells (don't cross from one row to another). The overexpanded cells only bend in one direction. Make sure the template is oriented so that the nome will be able to bend around the cylinder.

Layup Assembly:Shown below are two stages in the layup process. figure 2 shows the mandrel has been sanded and chemically treated. The first wrap around layer of release film has been added. In figure 3 below the first layer of carbon fiber has been added over the release film.

● put on powderless nitrile gloves● screw the coupling rings into the dummy rings● without touching the molding surface of the mandrel, slide the ring assemblies on to the

mandrel● screw the dummy rings to the mandrel● place the mandrel back on the mandrel holder● lay a layer of blue, perforated release film on the mandrel. It should span the entire

molding surface of the mandrel and not overlap with itself or the coupling rings (figure 2).

Figure 2: mandrel wrapped in releasetape

● lay the adhesive strips onto the lower adhering surface of each coupling ring. It may be necessary to use spray adhesive to get it to stay on the rings.

● lay the inner layer of carbon fiber onto the mandrel. It should overlap the adhesive strips completely, butting up flush against the step of the rings (figure 3). It should overlap itself by about an inch.

Figure 3: The first layer of carbon fiber added to the layup

● lay the inner adhesive layer over the carbon fiber. Rotating the mandrel while someone holds down the adhesive (wear powderless nitrile gloves) works well here.

10

● once the adhesive is well aligned, peel a section of it back. Someone should still be holding it down.

● apply a light coating of spray adhesive to the exposed carbon fiber● roll the adhesive back down onto the carbon fiber● roll back the remaining adhesive● apply spray adhesive to the newly exposed carbon fiber● roll the adhesive film back onto the carbon fiber. There should be no overlap of the

adhesive film with itself.● lay the overexpanded nomex honeycomb core over the mandrel● perform the same roll-back and spray technique to adhere the nomex to the adhesive

film● scrunch the overlapping portion of the nomex, so that it sits flat on the cylinder● use the roll-and-spray technique to adhere the outer layer of adhesive film to the nomex.

Take care to only apply the spray adhesive to the inner surface of the adhesive film. Applying spray adhesive directly to the nomex is useless. There should be no overlap of the outer adhesive film onto itself.

● use the roll-and-spray technique to apply the outer layer of carbon fiber to the outer layer of adhesive film. There should be about an inch of overlap between the carbon fiber and itself. Take care not to tug on the carbon fiber. When it’s in the oven any deformation to the carbon fiber cloth will relax, forming wrinkles.

Vacuum bagging:After adding the final layer of release film, the layup is ready to be vacuum bagged. Figure 4 shows the layup wrapped in it’s final layer of release film.

● lay the outer layer of blue, perforated release film onto the mandrel. It should have about an inch of self-overlap (figure 4).

● secure the release film with either kapton tape or flash tape● lay the breather material (“diaper”) over the mandrel● secure the breather material with either kapton tape or flash tape● form a 7 inch diameter tube of vacuum bag material by joining a sheet of it along one

edge using the yellow vacuum bag tape (“goop tape”). This tube should be at least 1.5 feet longer than the mandrel.

● gently pinch any air channels out of the goop tape● form a 5.5 inch diameter tube using the same method● stand the mandrel upright● slide the larger tube over the mandrel● slide the smaller tube into the mandrel● place the mandrel on its side, on the plastic surface of the work bench● adjust the bags so that they are centered around the mandrel● using the goop tape, seal one end of the inner tube to the corresponding end of the outer

tube● take care that there is no air channel formed by the overlap of the outer tube with itself● gently pinch any air bubbles or air channels out of the goop tape

11

● cover the end of the vacuum hose with diaper, securing it with kapton or flash tape● place the diaper end of the vacuum hose onto the dummy ring near the open end of the

bag● tape down the hose using kapton or flash tape● carefully apply goop tape to the vacuum hose, so that it may seal to the tubes● carefully apply a ring of goop tape to the inner surface of the outer tube● seal the inner and outer tubes to each other● gently pinch any air channels out of the goop tape, taking special care to make sure the

hose and tubes form a complete seal. Make sure there is no air channel formed by the overlap of the outer tube with itself.

Figure 4: layup wrapped in final layer of release tapeand ready to be vacuum bagged.

Curing:● set up the aluminum (not plastic) mandrel holders in the oven● thread the vacuum hose through the hole in the oven wall● gently place the mandrel on the aluminum mandrel holders in the oven, while someone

holds them in place● attach the outside end of the vacuum hose to the venturi pump● attach the venturi pump to a compressed air supply● while slowly turning on the compressed air supply, check the bag for hissing (this

indicates a leak which must be sealed with goop tape)● turn the air supply on fully● arrange the thermocouple to be directly underneath the mandrel● close and latch the oven● make sure the oven is plugged into two separate circuit breakers● make sure both relays on the side of the oven are in the “on” position● ramp up the the oven at 3 deg F per minute until it reaches 350 deg F● after it reaches 350 deg F, hold the oven at that temperature for two hours

12

● ramp down the oven temperature by 5 deg F per hour. Use the oven door to control this cooldown, opening it slightly for faster cooling.

Removal from mandrel:● once the oven has reached 100 deg F, open the door and slide vacuum hose into the

oven, through the hole in the oven wall, using insulated gloves● remove the mandrel from the oven using insulated gloves and place it on the mandrel

holders on the workbench● cut open and dispose of the vacuum bag● remove the hose, diaper, and release film from the mandrel and dispose of them● place four mandrel removal blocks on the floor● place the mandrel in the blocks, using insulated gloves● place a piece of wood on the top of the mandrel● tap the piece of wood with a hammer to release the mandrel from the module● with one person holding the mandrel with insulated gloves, gently pull and wiggle the

module off the mandrel in the direction of the male ring. Take care not to damage the module or mandrel. If necessary, run cold water through the inside of the mandrel. This will thermally shrink the mandrel, allowing it to slide off easier.

● place the mandrel back on the mandrel holders on the workbench● remove the blue perforated release film from the inside of the module

Results:We developed veins during the curing process. It is believed that this is the result of either the loose vacuum bagging material or the 3M AF30 scotch-weld adhesive, this is shown in Figure 5.

Figure 5: Final carbon fiber rocket module

Hypothesis:

13

Aluminum caul plating will be precisely measured and wrapped around the layup, between the release film and the breather material. Ideally this will distribute the pressure and effects of slack bagging material.

14

Nose Cone Design

Shape:The type of nose cone we selected was a Von Karman ogive. There are several types of nose cones that work to varying degrees depending on velocity of the rocket. The von Karman shape minimizes the wave drag at supersonic speeds, which account for the majority of the drag on the rocket.

Nose Tip:A mountable aluminum tip is planned for the end of the carbon fiber nose cone. The purpose of this is to absorb some of the heat generated at high velocities. The epoxy used in carbon fiber cannot handle temperatures above 175 degrees Celsius. If it crosses this temperature, it will begin breaking its cross-linked cross linked structure. In this way, the nose cone could incur invisible damage during flight.

To design this tip, we looked for two pieces of information: how long the metal tip needs to be and what radius it should have. The empirical equation I we from Peter, the NASA contact from the liquid engine team, is

q̇=α ρ1/2R−1/2v3 ,where q̇ is the heat generation rate, ρ is the air density, R is the tip radius, and v is the free stream velocity. This allows us to determine the necessary radius to put at the end of the nose cone to prevent the aluminum from melting.

We also need to find the temperature distribution over the nose cone. Initially, we tried to estimate the of the aluminum tip by comparing the wave drag to the total drag. However, when the wave drag was calculated, it was found that the theoretical wave drag on the nose cone was greater than the total drag predicted by the OpenRocket model.

Suspicious that there was a problem with the OpenRocket model, we compared the L13a model with the L12 model, which is known to be very accurate. Plotting drag coefficient as a function of Mach number (figure 6), L12 showed the characteristic bump in the drag coefficient in the transonic region. However, L13a does not show any transonic behavior, which should not be possible. So, we discovered that the OpenRocket model created by the 2014 capstone team must be refined.

15

Figure 6: strange lack of transonic behavior in the LV3 model.

We are currently seeking help from the liquid engine team to create a computational fluid model of the flow over the nose cone. This will give the distribution of the skin temperature on the nose cone. With that distribution, we can choose the length of the aluminum tip (plus some margin of safety), such that the carbon fiber only experiences skin temperatures below 175 degrees Celsius.

We are also working to get an electrical or mechanical sophomore or junior to set up an Arduino to act as a temperature sensing module for the rocket. By installing thermocouples in the nose tip, coupling rings, and motor mounting flange, we will be able to verify our design after the first flight. This would make all future launches safer, by alerting PSAS to any dangerous temperatures created by changes to the motor size and payload.

16