First Build Day of the Year

It was great to see everyone and great to see the hardware again. In the time that's elapsed since any fabrication efforts, a number of refinements have been identified. 

One of the most important solutions involved the redesign of the two valves and one actuator piston which release the kerosene and liquid oxygen for launch. 

The valve, which is a modification of two stainless steel ball valves bought off McMaster Carr, was designed off our old valve, which worked perfectly at room temperature, but froze up and failed to acctuate when tested with liquid nitrogen. The fix - which included better machining -  substituted the 1 inch diameter piston for a 1.5 inch diameter piston, while also allowing for us to deliver a higher actuation pressure – meaning that the valve levers will receive a much greater force, and hopefully open much more reliably.

After many conversations, we have realized that this valve design will always be inferior to a truly cryogenically designed valve - The only stumbling point being that we can't afford one. We hope to incorporate cryo-friendly valves into our Static Fire System refurbishment, and hopefully in future rockets. 

And of course, lots of hardware sorting. 

And of course, lots of hardware sorting. 

In addition to the valve switch up, the dewar that we fill our oxygen and nitrogen from will be higher pressure, meaning that the dribble  that was causing us problems previously will hopefully be increased to a more reasonable fill rate.

Finally, the composite aerodynamic structures will be receiving a refinishing, and all logos of our beloved sponsors will be bestowed on the rocket with vinyl stickers. 

 

New Recruits

With so many projects, we had to rustle up some potential new members yesterday at Engineers on the Green.   

John Marcozzi looks way too peppy for his senior year.  

John Marcozzi looks way too peppy for his senior year.  

The old TriD engine and the new Ignus engine served as talking points along with some electronics assemblies.  

The old TriD engine and the new Ignus engine served as talking points along with some electronics assemblies.  

Darren Charrier displays his recent advances in PVC Banner technology. 

Darren Charrier displays his recent advances in PVC Banner technology. 

New Static Fire System in the Works

We are planning to revamp our Static Fire System, and will be addressing troublesome aspects of our past system in the new design. 

Our last engine firing, in the YouTube video below, showed signs of an initially fuel-rich burn. The hard start could have resulted in a failure, but stabilized into a characteristic burn about half a second into the firing. 

There are several aspects that could have caused this hard start. One likely reason is that a previous scrubbed firing attempt, which depressurized and expelled the JP4 and Lox through the engine, might have left a pool of kerosene in the curved lower half of the combustion chamber - unknown to us - and eventually causing the fuel-rich plume when we recycled and fired again. 

This problem could have been averted with a Static Firing System that mounts the engine with a downward tilt (nozzle angled slightly down), which would allow any kerosene from a previous firing or any that comes out due to an unintended leak to drip out of the combustion chamber.

SEDS' current SFS now has 3 firings under its belt, but it's time for a few improvements.  

SEDS' current SFS now has 3 firings under its belt, but it's time for a few improvements.  

Another key issue is our tank fill levels. In the firing video, once the engine thrusting has ended, a large amount of kerosene can be seen gushing out of the engine. This is because our Lox tank was drastically under filled. The current method we have been using to fill our lox is a simple visual confirmation of venting, which should indicate the liquid has reached the maximum fill, but in practice this is much more difficult when you are out in the middle of the Mojave. 

Our refurbished SFS will hopefully have recording scales beneath both the fuel and oxidizer tanks to confirm fill levels in the most direct way possible. The fill tubing for the tanks can be replaced with flexible hoses and the weight of the tanks can be tared out. This will also give us a fuel and lox flow rate which we can compare to our already installed turbine flow meters.  

Progress with Vulcan I

There has been a number of refinements and fundamental changes to the recovery and electronics bay portions of the rocket in the past few weeks, so I've listed the main ones by category! It's going to be a monumental push to finish the project by our deadline

– which is only a week after the academic quarter ends at UCSD. With classes, internships, and occasional meals I suspect that the next month will present numerous late nights for all the brave students of SEDS who have committed themselves to this project.

The SEDS Kickstarter campaign, managed by the business team, not only surpasses it's $15,000 goal, but makes staff pics for Kickstarter as well as receiving shoutouts from numerous notable individuals in the aerospace industry. 

The SEDS Kickstarter campaign, managed by the business team, not only surpasses it's $15,000 goal, but makes staff pics for Kickstarter as well as receiving shoutouts from numerous notable individuals in the aerospace industry. 

While I had originally anticipated that the electronics system as a whole would be minimal, the advantage of having a large volume to work with has given us the opportunity to accommodate our own experimental parachute ejection system, which will be a custom PCB that we have designed ourselves. While I'm not an expert in electronics, one interesting aspect of this system will be it's constant communication through an antenna – mounted lengthwise in the nosecone – that will constantly relay the flight telemetry back to the ground. Additionally, the system will be recording the acceleration of the rocket, and perhaps even looking at the double integral of the acceleration to give live estimates of altitude in addition to the barometric altitude estimates that we will be recording using a commercial flight altimeter. This will be an entirely separate system, along with a separate power source for redundancy. 

Each parachute compartment – The main and the drogue – will contain two ejection wells, each with two ematches installed. Each parachute ejection system will be wired to one of the ematches in every single ejection well, meaning that while one system will inevitably fire first – even if only by a fraction of a second – we have complete redundancy in both the systems and the explosive charges.

Both the main and drogue parachutes attach to U bolts on the same bulkhead.  

Both the main and drogue parachutes attach to U bolts on the same bulkhead.  

Because of the space-conservative design of the system, the wires caring the current for parachute ejection must necessarily bridge the gap between the airframe and the shoulder of the nosecone, which must separate during parachute deployment. It is such that at this junction we must have a connector which is tight enough to not experience loss of continuity due to vibrations or launch acceleration, but loose enough so that it doesn't impede the separation of the nose cone and the airframe during parachute deployment. 

Because of this connection juncture, it is especially critical that the shear pins holding in the main parachute are not accidentally sheered during the deployment of the drogue parachute at apogee.

By consensus, we have also decided to omit the secondary (forward) bulkhead supporting the electronics sled tube against the inside of the nose cone that I had drawn in an earlier blog post.  This additional support will instead come from two-part expansion foam which will be poured into the cavity created inside the nosecone prior to the installation of the centering bulkhead at the aft of the nose cone. 

The bulkheads are created on AutoCAD, to be cut and etched on Lasercamm. The etching allows for outlining the placement of components and struts to reduce the risk of mistakes during assembly.  An etched index mark also allows for subsequent cuts to be made by replacing both the part and the negative plywood in the same position on the Lasercamm. 

The bulkheads are created on AutoCAD, to be cut and etched on Lasercamm. The etching allows for outlining the placement of components and struts to reduce the risk of mistakes during assembly.  An etched index mark also allows for subsequent cuts to be made by replacing both the part and the negative plywood in the same position on the Lasercamm. 

Vulcan 1 Nosecone and Electronics Bay

We have decided to store the electronics for the recovery portion of the rocket in the nose cone, to better make use of this space.

For the bulkheads in the nose cone, we won't need any aluminum brackets. This is because the bulkheads in the nosecone don't have to hold as much force – they are either only supporting the electronics bay sled or the weight of the nosecone itself, which is significantly less than the bulkheads below, in the main recovery portion, which will be holding the load of the entire rocket upon recovery.  

image.jpg

The drawings above provide a more in depth visualization of what the nosecone might look like. The number of bulkheads can vary depending on the configuration we decide on. In the drawing that I've made, there are two bulkheads just for the electronics sled portion (in addition to the actual mounting surface for the electronics), as well as two bulkheads that center the electronics sled tube in the nose cone. While technically only one bulkhead is needed to center  electronics sled in the nose cone, adding more bulkheads helps to support the fiberglass structure of the nosecone.

For the actual electronics bay sled,  I've included a two layer design in the illustration. This is helpful because it provides more surface area for mounting. Additionally, it allows us to store important components, such as the batteries, that need extra support in between the two layers, ensuring that they will not become dislodged during the flight. Because of the relatively large mass of batteries, they are prone to becoming dislodged by the acceleration of the rocket during lift off (or the jerk of the rocket due to recovery explosive charges). Obviously, this would constitute a destructive scenario for the rocket, because the electronics controlling recovery would cease to operate. Electronic components with less mass, such as circuit boards, terminal blocks, and LEDs can be stored on the outer layers of the electronics sled where they are also more accessible upon removal of the electronics sled. 

We have a great deal of freedom in the design of this nosecone. This is fantastic, because it allows us the liberty of having fun with what we do! If we get ahead of schedule, and have some spare change, there's nothing that can stop us from adding on features that will improve the overall experience of the rocket, such as cameras, or the airspeed sampling pitot tube set up that I included in the drawing.

Of course, there are still several crucial issues that must be tackled, such as how electronics will be armed (turned on) when the rocket is on the launchpad, or how the wiring for triggering the recovery charges will cross the interface between the nosecone shoulder and the recovery bay airframe tube. 

I'm excited to see how the design process continues as we recognize an account for the difficulties we will inevitably encounter in integrating the entire system with the recovery Bay as well as the rest of the rocket.

To a general scale - not all dimensions were known off hand.  

Concept Development for SEDS Vulcan-I Rocket

There's been a lot to hash out on this rocket design. One of the sections that can be started on immediately is the payload bay for the main parachute and drouge parachute, which constitutes a separate system from the fuel tanks and pneumatic lines in the lower section of the rocket. This recovery section will have an electronics bay, stored in the nosecone, which will house all the electronics needed for recovery in addition to telemetry and flight data electronics.  

 

Whiteboard sketches of the entire recovery bass section. 

Whiteboard sketches of the entire recovery bass section. 

There will be multiple leave aluminum ejection wells which will contain the Blackpowder charge and seal the electronic match for ignition needed to blow out the parachutes.  These will be installed on the same bulkhead that the recovery lines attached to via a stainless steel U bolt. 

 

The nosecone will be purchased commercially from Public Missiles LTD.  unfortunately, the standard for a nose cone and high powered rocketry around this size is 7.5 inches rather than 8 inches – this means that there will have to be a half inch shoulder transition on the lower half of the recovery bay in order to enlarge the diameter to the 8 inch diameter of the fuel tanks for the rocket. This will likely be done by applying a filler around the circumference of the recovery day, sanding it down to the proper dimensions, and adding a strip of fiberglass over it.

some initial sketches I made to better understand the layout and vision of UCSD's SEDS newest rocket project.  

Two layers of fiberglass with overlap for leading edge of the fin. it might be better in practice to do is to separate pieces rather than four separate pieces.

Two layers of fiberglass with overlap for leading edge of the fin. it might be better in practice to do is to separate pieces rather than four separate pieces.

General system overview. One key feature will be an external parasitic shroud covering up the pneumatic lines that run from the fuel an oxidizer tank the rocket engine.  The configuration and extent of the cup recovery system will have to be determined once the recoverable weight of the rocket is known, but I'm expecting around 4 foot drogue a 12 foot main parachute.

General system overview. One key feature will be an external parasitic shroud covering up the pneumatic lines that run from the fuel an oxidizer tank the rocket engine.  The configuration and extent of the cup recovery system will have to be determined once the recoverable weight of the rocket is known, but I'm expecting around 4 foot drogue a 12 foot main parachute.

A Weekend at FAR

Took a few days out of San Diego to visit the friends of amateur rocketry site in Mohave, California.

Took this awesome slow motion video of an I impulse APCP rocket lifting off.  

A beautiful sunset panorama of the FAR our side of the surrounding lake bed.

A beautiful sunset panorama of the FAR our side of the surrounding lake bed.

I've always liked this logo. 

I've always liked this logo. 

Picked something off the bookshelf to return next time I'm out. 

Picked something off the bookshelf to return next time I'm out. 

A finalized bill of materials for underwater camera chamber

image.jpg
image.jpg

Once these items are purchased we should be able to move forward with the design without any hiccups! The one issue I could foresee is the smoothness of the fiberglass sheeting I'm ordering off McMaster Carr… I hope it's smooth enough that it will have a good seal with the red rubber that were using as a gasket material.

Initial sketches for underwater positive pressure electronics chamber

The two closures on either side of the tube take advantage of the already present positive pressure to compress a gasket that seals the inside of the chamber.

The two closures on either side of the tube take advantage of the already present positive pressure to compress a gasket that seals the inside of the chamber.

An actual scale lateral view of the chamber. 

An actual scale lateral view of the chamber. 

The current design has a cylindrical functional space of 8 inches and 2 inches in radius. I will look in the calculating what wall thickness is needed for the tube for a given depth, and at what depth we can expect to see structural failure given whatever wall thickness we decide on. 

The 3/8 inch thickness of both of the end plates is in part because I want to be able to secure things to them by tapping partially through without breaking the seal to the outside.  I want to include four threaded rods placed in a square pattern and circumscribed by the circle of the inside of the tube. This will allow for easy mounting of electronics at any depth within the tube.

Next, I want to create frontal views of both the gaskets and the two endplates. Hopefully, we will be able to cut the gasket material using some sort of laser cutting device. Perhaps the laser cam at the undergraduate design lab? We will have to see about this after meeting hopefully sometime early next week. 

This is old, but I forgot how good of a job I did one it. I need to do something like this again for this quarter. 

Acrylic Prototype

I finally got around to cutting an acrylic prototype with a few major changes. Firstly, the wood portion has been omitted entirely. This is mainly because I want to focus on completing the project rather than aiming for a project that is too time consumptive to realistically create in the span of a quarter. 

Secondly, the electronics package under the board was finalized, and an etched grip pattern was created as a separate layer.  

The grip is etched into the prototype board as a hexagonal pattern. 

The grip is etched into the prototype board as a hexagonal pattern. 

The cut was very clean, with the holes for the trucks and Arduino lining up perfectly with the actual mounting holes for the parts. The bolt holes, however, we're consistently around 5-10 thou too small in diameter to accommodate the bolts with a close fit. I reamed the holes out on the board to test fit the components, but will increase the hole size accordingly for the actual polycarbonate board. 

The acrylic prototype board with one of the trucks attached. 

The acrylic prototype board with one of the trucks attached. 

I greatly enjoy the aesthetics the shape and etching provide, but am concerned that the design has not had enough revisions to be finalized. My hope is that by having a physical model, I will not only be able to more easily craft the struts and support paneling (and come up with a plan for the electrical wiring), but also have a chance to reconsider the spacial layout and aesthetic appeal of the board's design as a whole.

One of my goals was not only to create a functional item, but to work with the physical properties of the materials, such as the photo-permeability of the acrylic and reflectivity of the aluminum and trucks, to create an object that is visually intriguing. Unfortunately, with the removal of the wood from the design, one of the most interesting artistic components - the interaction between the complex, organic texture of the wood with the homogenous and unnatural surfaces of the acrylic and metal - is lost. I hope to regain this dynamic by incorporating a natural texture somewhere else. I haven't decided what material, but would love to return to some sort of hardwood, as I originally intended. 

My next step in the project will be fabricating the struts and panneling, and testing for deflection under a load!  

Longboard Design Moves to AutoCAD

I've transferred my on-paper thoughts to an actual design on AutoCAD. I hope to finalize the design soon, but there are a few concerns....

The first, main concern is accurately predicting the deflection of both the aluminum struts and the acrylic. I have explored the math on the issue, but it starts to look a lot more like an estimate when complicated shapes are involved. The moment of inertia of the acrylic (shown in white) is troublesome to calculate by hand. All of this makes me feel that a Solidworks file should be made for the design, so that I can do some actual modeling for different loads. 

The second concern is the differences in flexibility between the wood handle portion of the board and the acrylic. While they are connected by the aluminum struts, I am concerned that when the board flexes under its load the acrylic will flex substantially more than the wood, resulting in strain on the struts and a possible uneven surface between the joint of the wood and acrylic on the top surface of the board. 

The simple solution should just be to omit the wood, as it is purely aesthetic, however one of the reasons I wanted to incorporate the wood was to create an intriguing material contrast between the man-made materials of acrylic and aluminum and the natural texture of the wood. I have enough acrylic stock to make two versions of the board, one with the wooden handles (shown in purple), and one that is entirely acrylic. 

My final predicament is the grip on the board. I had intended on creating a gripped surface by using the etching function of the lasercamm, but I an concerned that this may not suffice. I hope to etch a test surface and see how well it grips compared to the conventional adhesive grip used on other boards. I want to ensure that the grip does not interfere with the aesthetic of the board, as an appealing final product has been one of my goals since I started this project. 

While these are all great considerations, I am tempted to simply make the board and skip to testing. The more I think about it, the more I see that my reservations of what might go wrong will only increase, and they will only delay me starting construction. There will always be uncertainty as to whether the design will come to complete fruition as I have envisioned, so perhaps the best approach is to simply produce the board and see what happens! 

Acrylic Longboard Attempt 2

After my first failed attempt at designing my own laser cut acrylic longboard, I've decided to try anew with a new resign that focuses on rigidity and durability. 

The redesign features a below-deck aluminum frame that adds support as well as hardwood features for asthetics.  

I'm not sure how far the project will actually progress, but I've already obtained the acrylic and the design has become surprisingly refined after only two weeks of in-class doodling. The next logical steps seems to be generating an Autocad of the peices to be cut and etched. 

 

image.jpg

Vapor Smoothing: First Test

I arranged a small test today to gain some practical experience in smoothing some spare 3D printed ABS parts. The setup was simple: 

  1. a 100 Watt heater (which had been sealed into a bore in a chunk of copper) was attached to an aluminum plate with some thermal paste
  2. A small, glass mason jar with around 5 mm of acetone in the bottom was placed on the aluminum plate to be vaporized 
  3. A spare ABS part was lowered into the jar with some fine steel wire for smoothing

Here's the general setup. I apologize for the photos all being screen shots from my snapchat - unfortunately I neglected taking actual photos and had to salvage these after. 

As shown, the aluminum plate heated up to around 80 C within around 10 minutes, and the vaporized acetone, visible by a slight condensation on the sides of the mason jar, filled to the brim of the container within five minutes. 

I dipped the part into the vapor and immediately saw the part begin to glisten - as though wet with enamel. The glistening continued all over the part until the gridlike texture of the 3D printed ABS completely disappeared. This only took around 40 - 50 seconds. The difference was quite visible, as shown by the two before and after photos. 

The second part that was dipped (not pictured) was dipped in several intervals with a swishing/stirring motion to agitate the vapor and evenly effect the part. 

The next step is to create a permanent setup that can accommodate larger parts and that has a feedback system using a thermocouple and a temperature switch to prevent the acetone from boiling over. The heater was hard to control, having only a laser thermometer the monitor the temperature, and twice the acetone in the bottom of the glass jar began to boil. 

I ordered a 2.5 gallon glass jar to smooth larger parts in and will hopefully construct a functional semi-permanent setup in the next few weeks. 

ABS Smoothing Bath Using Vaporized Acetone

The first useful page

  • Jar placed on heating plate
  • 3-4 mm of acetone in the bottom of the jar
  • Heat plate to 120 C initially 
  • Once the jar is visibly filled with vapor, drop plate temperature to 90 C for treatment 
  • Smoothing takes several hours
  • Once smoothed, place a lid on the jar and cool the plate, condensing and conserving acetone 
  • Allow part to sit for 10 minutes while residual acetone evaporates 

The second useful page: 

  • 70 - 90 C for plate
  • Sealed metal container
  • Fan inside to "ensure uniformity of treatment" 
  • 10 minute bathing period 

The third page: 

  • Rice cooker is used
  • Custom pin board goes into bottom of heating vessel to allow for part placement while smoothing 
  • Part is cooled in fridge beforehand to ensure acetone condenses onto the surface
  • Smoothing takes several minutes 
  • If part gets too hot, condensed acetone will boil off and create pit marks in surface 
  • Test if filament is pure ABS by dissolving a small clipping completely in acetone beforehand 
  • Part must be printed from pure, high-quality ABS to smooth well