UB SEDS

Overview

The University at Buffalo Students for the Exploration and Development of Space (UB SEDS), is an engineering student organization dedicated to promoting the development and exploration of space through engineering projects and education. SEDS is a national organization with chapters all around the country.

Our chapter generally has a number of project and educational groups related to rocketry, astronomy, and robotics.

In terms of rocketry, there are two groups. The first of these is High Power, which certifies individuals to fly level 1 and level 2 rockets, students start with raw materials and build and fly their own rockets. The second of these groups is the IREC team, which is what I have predominantly participated in. IREC stands for the Intercollegiate Rocket Engineering Competition, which is hosted annually as part of the Space Port America Cup. This is an international competition, in which teams from all of the world come to compete. There are several categories, but during my time in SEDS we have always launched in the 10k SRAD category, meaning a target altitude of 10,000 ft, using a student research and developed solid rocket motor.

As for the other groups, our astronomy group performs analysis of publicly available astronomical data, and also runs a book club, and we have a shared battle bot team with UB's AIAA chapter.

My experience with UB SEDS began in the fall of 2018 when I first enrolled at UB. I spent that year performing various tasks with the structures sub-group of the IREC team, and went to the 3rd Annual Spaceport America Cup in Las Cruces, New Mexico later that year. Our rocket that year, PEDEX, had a successful flight, and I came away with an incredible experience. The following academic year, I moved over to the propulsion hardware sub-group where I would find my permanent position. That year I first learned how to machine in the machine shop, and fabricated various components of our SRAD motor, as well as working with my lead on the structural analysis of the motor. That year was unfortunately cut short by the arrival COVID-19, and I took over as the propulsion hardware lead the following year, still in the midst of a pandemic. As the propulsion hardware lead, I have learned an unbelievable amount about how solid rocket motors are designed, analyzed, simulated, manufactured, and tested. My experience as a lead on this sub-group has gone so far beyond what I was exposed to academically, that I would classify it as the single most significant educational experience of my undergraduate career. In an effort to record everything I've learned, and in order to help the club move forward with more ambitious projects, I've compiled a manual which comprehensively describes the process and theory behind developing a solid rocket motor. I hope that this streamlines the learning process for those who come after me. I've also completed a propulsion script in MATLAB, which works to optimize the design of our SRAD motors, and then simulate their performance during a burn as a function of time.

On this page, I will take you through absolutely everything I have done with UB SEDS up to this point.

Composite Materials Testing

So during my first year with the structures sub-group, I mostly made small contributions like cutting and sanding fiberglass. I also sank many hours into trying to learn finite element analysis (FEA) for composites in Femap NX Nastran. I never quite got a handle on that software, but it did set me up to learn FEA in ANSYS a year later. Now, one thing I did do that year that I am proud of was participate in the tensile and compressive testing of our fiberglass. Since we lay it ourselves, we have to determine its unique mechanical properties to ensure it can withstand the forces of flight. This was why I was doing this in conjunction with learning Nastran, so that I may perform FEA on a model of airframe.

So, I helped lay up some fiberglass tabs, and we sent them to our materials lab where we complete our materials course for tensile testing. I took all the data files and processed this into stress/strain curves using Excel. I obtained Young's Modulus for all three normal directions relative to the weave of the glass, and the yield and fracture strengths of the fiberglass parallel to the weave. All of this made it into our final report for IREC that year. Unfortunately, we did not have appropriate strain gages to examine the shear modulus of our composite. So, when I would eventually learn how to analyze composites using the ACP library in ANSYS, I had to ballpark this property using a value within the typical range for fiberglass.

Using what we did have, we found that our airframe had some ridiculous factor of safety for the anticipated aerodynamic forces we would be subjecting it to.

As an aside, I also learned how to mix and cast our solid propellant this year. This was my introduction to the propulsion chemistry and propulsion hardware sub-groups.

Third Annual Spaceport America Cup

Our team finished PEDEX, and that June we drove down to Las Cruces over the course of several days. I cannot capture in words how incredible the entire trip was. I had never seen so much of the country before. We stopped in several cities and saw some of the major attractions, tried all the food, and split all the driving. When we arrived in Las Cruces, we began preparing for the conference and the competition itself.

At the conference I got to listen to the speakers talk about the competition, the future of the space industry, and staying safe in the desert. I walked around during the poster session and spoke to students from all around the world about their rockets and payloads. Then I walked around the sponsor booths and spoke to representatives from Virgin Orbit, Blue Origin, and several other space industry names. These interactions were some of the best parts of the whole experience for me, and really drove home that I was in the right place.

The next few days we were in the desert prepping for launch and watching ninety or so rockets take flight. There were some failures, some spectacular failures, and a number of successful flights. When PEDEX went to the launch rails with the team leads, I waited patiently with the rest of the team and our advisor. When the time finally came to launch I watched as PEDEX shot straight into the clear desert sky, sputtered a bit, and continued straight up to over 8,000 ft.

Ultimately, we didn't place in our category, but we were more than satisfied with our successful flight nonetheless. The rest of our time in New Mexico was spent seeing the sights. At one point we visited the White Sands Missile Museum, and there I got to see everything from a V-2 to the Talos missile.

On the way back, we celebrated our successful launch and saw more sites. PEDEX now hangs in Davis Hall not too far away from the shared SEDS/AIAA lab.

Propulsion Hardware Lead Summary

As I mentioned before, as the propulsion hardware lead I performed every step in the development of our SRAD motor. I worked on everything including the design, analysis, simulation, fabrication, and testing of our motor. Not only this, but I learned all the theory behind the motor's operation, and the nuances behind different design decisions. To capture all of this information and make it easier for other students to learn, I've compiled all of my knowledge into a comprehensive manual. Additionally, I put together a MATLAB script which streamlines the design of the motor and allows users to instantly simulate the burn using that same configuration.

In the following sections, I'll go into detail about all of this.

SRAD Motor Overview

Our SRAD motor has a fairly typical configuration. The largest component of the motor, which is the first we design and fabricate, is an aluminum 6061-T6 casing. Next we have our forward closures, which are essentially plugs used to seal the top end of the casing. These are also aluminum 6061-T6, and are retained using an inverted retaining ring. For the other end of the casing we have our nozzle. The nozzle is a simple conical converging-diverging nozzle made of isomolded graphite. The nozzle and the forward closure each use two o-rings in series to create their static radial seals.

The propellant used in our motor is cast directly into a paper phenolic liner. This insulates the motor casing from the high temperatures of combustion. As the paper liner burns it carries heat away with it, this is an ablating insulator. The grains themselves are simple BATES grains, which are long cylinders with a hollow core that acts as the initial burning surface. BATES grains are almost the easiest the fabricate, and also almost the easiest to simulate.

SRAD Motor Development

The course of developing our motor can be described in the following steps, which are split between our propulsion chemistry and propulsion hardware sub-groups.

Propellant Characterization

Small Scale Test Fires
Pressurized Strand Burns

Motor Design and Preliminary Simulations
Motor Analysis

ANSYS
By Hand

Motor Fabrication
Forward Closure Design & Fabrication
Hydrostatic Testing
Nozzle Design and Fabrication
Test Firing
Data Processing and Validation

You'll notice later that my manual does not quite follow these steps in order, because we have some small 38mm motor casings we use for propellant characterization that require some concepts to be taught in advance in order to operate. However, the order listed here does indeed accurately reflect the development of the full scale 98mm motor.

Motor Analysis

As we design our motor casing, we must perform analysis to ensure it can withstand the anticipated operating pressures. We analyze the casing using a model in ANSYS. This involves preparing the geometry, setting up materials, setting appropriate contacts between components, creating a good quality mesh, applying appropriate boundary conditions, understanding what failure theories to apply, and performing a convergence study.

It took me a long time to understand how to create our current model, and I learned a lot of different ways to assess model quality, interpret results and mitigate singularities. I've recorded everything I know about ANSYS and FEA in general in chapter 8 of my manual.

The end result of my model is something that produces realistic results that exhibit trends we can observe in some of our older casings that have actually failed in some way. Coupled with a robust by hand calculation of the stresses in the motor casing, we can be confident that our model is returning reliable information.

It is my intention to have simulate a number of different geometries in terms of thickness and edge margin in order to create a statistical catalog to aid in designing motor casings in the future. The information in this catalog would be organized according to dimension, and maximum allowable pressure. Maximum allowable pressure in this case being what the casing can be safely hydrostatically tested to.

Motor Simulation

To simulate the burn of our motor to obtain useful information such as the chamber pressure and thrust as a function of time, total impulse, and other parameters, we can use existing software like BurnSim and OpenMotor. These programs take motor geometry, grain geometry, propellant characteristics, and nozzle geometry, and output the above parameters.

It's useful to perform these simulations at the beginning of the design process to ensure that we can meet our performance requirements, and to monitor the maximum expected chamber pressure for structural concerns.

We always perform simulations for any given configuration before a test fire so that we may validate this theoretical data with hard experimental data from our test stands.

As part of my MATLAB propulsion script, I've written a similar simulation to BurnSim and OpenMotor. As a result, I understand how these softwares work fairly intimately. The burn is always simulated as a function of the regression depth into the propellant grains as they burn away.

Motor Casing Fabrication

Fabricating the casing is the most time-consuming portion of our work in the machine shop. We start by turning down the external diameter of our casing on the lathe. This is extremely time consuming and carbide cutting tools should be used to speed up the process.

After spending an entire work day making the long external cuts, we come back to deal with the inside. Because we require center support, we have to move out of the student machine shop and into the main machine shop. In here we bore out the ends of the casing to create a concentric bore, and then continue to bore until we reach our bore diameter of 3.512 inches. While we're at it, we have to bore out the internal retaining ring grooves.

Finally we have to return to the machine shop for a final day to create an external retaining ring groove on the outside diameter of the motor.

Forward Closure Fabrication

Machining the forward closure is very straight forward and can be done entirely in the student machine shop, and is the only component we machine on both the lathe and the mill.

We start on the lathe by facing both ends, turning down to our fit external diameter, and putting in the O-ring grooves. Next, if the current forward closure has a center hole, like the test fire forward closure, we'd drill and tap that on the lathe. Otherwise we jump on the mill.

On the mill we can start by drilling and tapping the holes for plumbing on the hydrostatic forward closure. All of our plumbing is threaded for 1/4 NPT. On all of the forward closures, we have to drill and tap blind holes for the eyebolts we use to disassemble the motor casing. These are threaded for 3/8-16.

Nozzle Fabrication

Like every other part of the motor, the nozzle is predominantly made on the lathe. Unlike every other part of the motor, the nozzle is made of graphite. This complicates things.

Graphite is extremely messy to work with. The entire time I'm machining it I have to hold a vacuum cleaner up to the cutting tool to try to catch a decent amount of the dust being ejected in every direction. For the smaller 38mm nozzles, one person can usually manage, but a full scale 98mm nozzle is a two person job.

We perform the usual operations, facing both ends, turning it to its fit external diameter, putting in the O-ring grooves, drilling out the throat, and boring out the converging and diverging cones. When we're finished, we quickly polish the nozzle using increasingly high grit sand paper, attempting to round out any sharp edges to improve the nozzles strength and resistance to erosion.

I don't actually have a picture of a 98mm nozzle in the shop so the image on the right is of a rough 38mm one.

Motor Hydrostatic Testing

After we have all the components of our motor fabricated, we have to certify it to fly at our maximum expected operating or chamber pressure (MEOP). To do this, we have to pressurize the motor up to 1.5X our MEOP for 2X our burn time. We accomplish this by filling the motor with water and pressurizing it using nitrogen gas.

During my time as hardware lead, I decided to certify a motor for 800 psi, which required pressurizing it to 1,200 psi. Since I didn't have a good idea of our anticipated burn time when I ran this test, I kept the motor pressurized for 30 seconds which is 5X our previous burn time with PEDEX. When I could establish a burn time, it ended up being approximately 7 seconds.

It actually took me 3 tries to get the test to run. I had issues with the plumbing and the assembly of the motor. Eventually after a little bit of research, some problem solving, and some trips to the machine shop, I got everything ready and the test went off without a hitch.

SRAD Motor Test Firing

Once everything has been rated for our MEOP, we can move immediately to test firing our full-scale motor. The chemistry sub-group will mix and cast our propellant grains in its phenolic liner, and we'll assemble the motor together. We have a large vertical test stand, which fires the motor towards the sky. A force transducer sits at the bottom of the stand and is calibrated to record the thrust throughout the burn. We use this data to validate our simulations, and our dynamics sub-group uses the data to adjust its estimations.

My Propulsion Manual

I learned A LOT in two years in and leading the propulsion hardware sub-group. It was not easy, I rarely had any good sense of direction, and I often found myself lost and confused. However, I do think that I've finally come to master this material. Everything from the nuances of using different tooling on the lathe, to the theory behind the motor's operation, to even the finest details of a successful model in ANSYS. I have learned, and overcome it all, but it was never easy.

It is for this reason that I decided to set out on writing my manual. I want those who come after me to be able to pick up where I left off, and take this team to the next level. I don't want anyone feeling like they're spinning their tires and going nowhere while trying to learn the basics as I had felt for so long.

My manual now sits at over 100 pages in length, but it is still a work in progress. To say the least, it is comprehensive. I more or less wrote a textbook. You can check it out online here if you have a google account. The contents are as follows.

Introduction
Web Resources
Software
Measuring
Machining
Fits & Interferences
O-Rings
Intro to FEA with ANSYS
Forward Closure Design and Fabrication
Design Prerequisit Parameters & 38mm Tests
Nozzle Theory, Design, & Fabrication
The Full Picture
Retaining Rings and Thrust Washers
98mm Casing Design & Fabrication
Hydrostatic Testing
Motor Assembly
Test Fire Preperation
Test Fire
Data Processing

My Propulsion MATLAB Script

Upon finalizing my journey through the theory behind our solid propellant motor, I decided I was ready to write a script which accomplished two things.

Optimize the geometry of the nozzle to achieve the maximum total impulse possible within the limits of an input motor geometry and rated chamber pressure.
Simulate the burn of the motor, returning plots of the chamber pressure, thrust, and other parameters as functions of time.

I achieved the first of these goals by iterating through every nozzle throat diameter within a range of realistic values, from smallest to largest. This is because the chamber pressure is in part an inverse function of the nozzle throat area, and so I can settle on the first throat area which does not create a chamber pressure exceeding that which we rated the motor casing for. This maximizes the total impulse of the motor because the impulse is as much a function of the chamber pressure as it is of the burn time. This part of the script will also return an exit area for the nozzle, such that it will expand the maximum chamber pressure down to the local atmospheric pressure.

The second goal was achieved by performing a linear regression of the propellant grains as they burn. Everything that defines the chamber pressure and thrust is constant during the burn except for the Kn, which is the ratio of burning surface area in the motor to the throat area of the nozzle. As the grains burn away, the burning surface increases for a time and the decreases. I use the burn rate estimated at every interval of the regression depth to determine the corresponding time step, and sum all of these to determine the burn time. The end product is quite similar to comparable simulation softwares, such as BurnSim and OpenMotor.

The final version of this script, also outputs a number of other parameters for any given simulation that would be useful to a designer, and includes a third mode, developed by our propulsion chemistry lead, which helps establish amounts of each ingredient of the propellant before a given mix.

You can view or download my script here.

On the right, I have a video demonstrating the script in operation.

Parametric Design

This is a skill that I picked up at my yearlong internship with Smart Walls Construction that I saw had a lot of potential for SEDS. Parametric Design in this context refers to parameterizing the geometry of a design, so that it can be driven from a spreadsheet. In my case, my models are automatically generated in Autodesk Inventor, and are driven by an Excel sheet. The user updates the designs parameters in the excel sheet as necessary, save, jump to the part or assembly in Inventor, hit update, and have the new geometry ready to go. This is a massive time saver!

The video I have on the right here demonstrates my current setup.

Utilizing VBA, you can push the capabilities of this so much farther. I would love to adopt my MATLAB propulsion script to VBA, and integrate it into this workbook at some point. The idea being you would enter your usual inputs and be able to generate parameters for an optimized nozzle, as well as simulate the burn of that motor. Parametric Design with CAD and Simulation. It would be super cool.

4th Annual Spaceport America Cup

With COVID-19, the original 4th Annual Spaceport America Cup was canceled. So, a year later, and still without a fully completed rocket, we got to attend the virtual 4th Annual Spaceport America Cup.

I'll admit I wasn't really excited about this, we were still working on our rocket, and like every other team, the only thing we could submit was our documentation. There were some really interesting workshops, some great speakers including former astronaut Mike Mullane. I also attended Blue Origins booth for the second time, which was actually very enjoyable. The representatives were openly discussing O'Neill Cylinders and other space-borne colonies. I kind of just sat there amazed that this was even on the table within my lifetime. Toward the end of the conference I attended the general lounges, spoke to students from as nearby as Waterloo and as far away as Brazil, and joined the unofficial Spaceport America Cup Discord server.

Overall I had a pretty good time, so it was a nice surprise when we got the news that our team won our category in the competition. Myself and the other leads thought it might be a mistake at first, but it turns out our documentation was enough to give us a win in the 10k SRAD category.

Of course we're still going to complete our rocket and launch it at some point in the near future, but this experience me some reassurance that we had done good work up until that point.

UB SEDS Logo

As a small note, one of the last things I did was design a handful of logos for our chapter. I have no graphical design experience, and I used an unrefined free software to do this, but I'm honestly really happy with the way they all came out. I don't know if these will actually be fully adopted, but I'm confident these might find themselves on a t-shirt during our next fundraising event.

Looking Forward

In the future I'll certainly stick around the Discord server to advise future students, and since I'll be a graduate student at UB for a while, I'm sure I'll pop in to help out from time to time with basic tasks. But, I'm still not done learning. I've taken up an interest in controls, and so for the next several years, I might be experimenting with thrust vector controls and avionics. This has become a popular subject in online rocketry recently and I think I can pull it off. We'll see! As for other ideas, I would like to prototype an airbrake, I saw a number of rockets with these in 2019, and who knows maybe I'll expand my propulsion script to include dynamics.

Whatever I do next, I'm sure it'll be fun and it will only increase my chances of breaking into the space industry. The sky isn't the limit I guess?