At this point, my patience for this project was starting to fade, I decided to go back to basics, and use soley clay for the nozzle, as I had seen people use online. I also opted to use the same translucent PETG filament for this test, as I thought that it would give a good view of the combustion so that I could see if there were any problems with inconsistencies in the burn. 

This motor was not my finest work, as the clay of the nozzle had cracked at the sides before launch, but it survived suprisingly well. It was a shame that it burnt through on the side of the motor, just as the highest rate of combustion was achieved, however this was to be expected as a similar phenomena has been observed in the past when using PETG filament.

Further, when drilling the hole down the middle of the filament, I encountered a problem. I usually hold the motor still between by legs so that I can drill directly downwards with my hand drill. This usually works well when fins are present, however due to the lack of fins, the motor would spin, making it very hard to drill a perfectly straight hole. Eventually, I gave up early, as it was near impossible to keep drilling without the assistance of someone else, and decided to commit to the test fire without the full length central bore. 

This may have been the reason that the motor faired so well, because the chamber pressure was lower than it would have been if all of the propellant combusted as the same time. Regardless, the clay worked well, and so I decided to return to basics for the foreseable future

This rocket is the first to be designed to the scale that I intended when I began this project.

When printing the nose cone, I encountered a number of problems. The first version, I printed in the horizontal position, which caused a number of issues. Firstly, it meant that the nose was filled with supports, which actually ended up breaking during the printing process anyway. This lead to the nose being filled entirely with broken support, which then rattled around. I feared that this would cause instability within the rocket during flight.

I then printed another version. This one did not have a barrier between the thread section and the rest of the hollow fuselage, meaning that I could then print vertically and pull out the supports rather easily.

Unfortunately, this rocket also suffered the terrible instability which had plagued all of the previous flights.

The rocket violently veered off course, colliding with the ground, causing the nose cone to be severed at the thread of the motor. This was becuase I hadn't printed the motor with a great enough sparse infill density, and therefore it wasn't strong enough. The motor proceeded to fly into the woods for a good distance, before being recovered by me and my dad from the foliage.

The second half of this flight was as close as we have come to stable flight!

The nozzle was mostly intact, however, the clay layer had vanished, likely having been ejected during flight. This throat was also heavily damaged, having been expanded by the exhaust. I fear that the offset of thrust was either due to one side of the motor igniting first, or due to a slight angle that the hole had been drilled at.

Eventually, I settled on this design, combining elements from the previous two iterations.

The rocket is partially enclosed, allowing for support during the launch. Despite this, the amount of material needed was still significantly less than what the first design required. There was no material surrounding the fins of the rocket, allowing for less resistance during the launch. 

Further, I allowed slightly more clearance than I designed into the base plate, as I realised that the base plate had slightly too snug of a fit, leading to concerns that the rocket may not be able to free itself, getting stuck in place.

The design elevated the nozzle off the floor, preventing exhaust gasses from damaging the exterior of the nozzle and the bottom of the launch pad. 

The supporting towers were connected to the outer base plate, leading to greater structural performance, and an increased ability to resist any outwards force exerted by the rocket during take-off.

I was very happy with the design of this launch pad, from both a functional and aesthetic perspective. It resembles the launch pad that the Russian Soyuz launches from. 

These test fires gathered information on the new nozzle design. This nozzle had a longer throat, with a sturdier design. The extra wall thickness was in an attempt to prevent a nozzle burn through, as observed on previous tests.

I also tested two different materials with this nozzle. I used a standard PLA motor as a control variable, and also used a motor made entirely of PETG - the same material used in the test fire stand.

I also added ridges where the fins would be located, simply by removing wing material outside of a certain diameter in my CAD software. This would give the motors greater structural rigidity, more accurately modelling the circumstances in which the motors would usually function. 

The standard PLA casing and nozzle functioned exceptionally well, even having enough thrust to slightly move the brick acting as ballast to the test stand. 

The PETG casing did not function quite as well, with it visibly expanding outwards from the pressure within the combustion chamber.

Both motors consisted of an updated nozzle, coupled with a layer of bentonite clay.

The large fillet, leading into the nozzle was intended to reduce the effect of corrosion on the nozzle by guiding the exhaust fumes towards the throat, comparable to the conventional solid rocket motors that this nozzle was based upon.

I neglected to realise however, that in this instance, it would only create an empty space between the clay and the nozzle, creating a cavity in which the hot plasma and gas could enter. This increased the risk of a high pressure area forming between the clay and the nozzle, possibly leading to a burn out on either side of the nozzle. 

Regardless, the PLA motor functioned well enough for me to immediately leap to the next step and implement it onto a rocket.

I conducted this next test as a control variable. I used the exact same setup as the carbon fibre motor with the clay layer, but printed entirely standard PLA matte filament.

This test did not have the same issue with burn through, verifying the benefit of the standard PLA. In this test however, another limiting factor became apparent. The plasma began to burn through the side of the nozzle bell, causing a loss of chamber pressure.

Further, the throat of the nozzle was almost immediately destroyed by the exhaust gasses, leading to an offset of thrust.

This motor casing did fair up much better in comparison to those made of the carbon reinforced filament, proving my hypothesis that the standard PLA filament is more well suited to this application.

After this test, I designed a new nozzle, with more material joining the nozzle to the thread. This was in an effort to prevent the burn through that I witnessed here.

This was a rather successful burn, producing quite a lot of thrust, and combusting quickly and cleanly. 

After the disappointing failure of my previous launch, I realised that I must first test the performance of the carbon fibre composite in my test stand. From the previous launch it was obvious that there was a problem with the plasma burning through the casing. In this test I aimed to determine the effect of the bentonite clay layer present between the propellant and nozzle, whilst also gathering information on the performance of this filament under high temperatures and pressures. I also used this to test out a new nozzle design which I hoped would increase the durability of the section and thus increase the thrust.

For both of these tests I used a nozzle with a larger bell that I had designed. This was after watching a number of engine design. I learned that the nozzle bell helps to expand the exhaust gasses, allowing them to flow at rates faster than the speed of sound. I have later realised however that this is mostly redundant due to the fact that the exhaust gasses in this instance are not flowing at speeds great enough for this to make a difference.

In the top video, I did not include the bentonite clay layer, instead relying soley on a plastic throat to channel the exhaust gasses. In the bottom video, I did include the layer of clay. 

From these two videos, it became clear that the clay layer did in fact make a positive difference to the performance of the motor. I believe this to be due to the fact that as the clay is exposed to the heat, it hardens, increasing durability. On the contrary, the purely plastic arangment rapidly deforms and is expelled, leading to a dramatic loss of chamber pressure. This can be seen from the amount by which the motor mount is pushed into the stand. The version with the clay even had enough power to slightly move the brick which was used as a support.

In both tests however, the carbon fibre casing burnt through  before the end of the burn.

I had initially thought that the extra rigidity of the material would be of benefit to the motor casing, however I believe that this is the property which caused the burn through of the casing wall. The original PLA filament deformed significantly during combustion, with the walls expanding outwards as the chamber pressure increased. It seems that from these videos, instead of becoming malleable under the temperature and pressure, the carbon filament instead became brittle, and fractured. The cracks in the material then propigated, allowing for the plasma to leave through the walls. The difference in deformation can be seen especially when comparing these tests to the next two tests, in which I used PETG. The PETG very obviously deforms, which may have allowed it to contain the high pressure combustion gasses without rupturing.

From this test however, I have determined that the clay layer is particularly important. this clay layer clearly reinforces the throat, allowing for greater thrust.

I will for these reasons not be using the carbon fibre reinforced PLA filament in future motor casings.

I believe this to be the first motor in which I used the part which screws onto the thread for stability during preperation. I will call this part the "Hammer Brace" - in the fact that it helps to evenly distribute the vertical force across the base of the motor, when I am compacting the propellant. 

I realise that I have never spoken much about the preparation of my motors - my technique has stayed mostly the same since the beginning of this project. I use a wooden dowel with a hammer to compact propellant and clay into a a casing. I then use a drill bit to bore out a centre core, with the aid of the nozzle as a template for drilling. For this launch I printed the aforementioned "Hammer Brace" which allowed me to keep the motor more stable whilst hammering. I believe that this helps to provide a consistent propellant density throughout the motor.

I had purchased a reinforced steel nozzle for my Bambu Labs P1S printer, which allowed me to print in reinforced carbon fibre PLA filament. This filament was advertised as having a higher temperature resistance and greater performance under stress, however this casing did not function as I had hoped. It became very apparent that the carbon fibre  did not fair up well to the pressure, as the plasma almost immediately burnt through the side, causing a dramatic offset of thrust.

The result of this "flight" was a rather sorry sight, with the gap between the nozzle and the rest of the casing having completely burnt through. I was not convinced that it was the material that caused these dramatic effects - I was still sure that the higher temperature and stress resistances would do nothing but good for my motor casings. For this reason, I went on to test this filament on future casings for static fires.

In this test, I attempted to assess the impact of two variables - something that I understand should rarely be done when attempting to gather meaningful data. I had two motor casings, alongside two nozzles; one of these nozzles had a significantly longer thread. My reasoning behind this was that a longer thread would provide a larger contact surface area between the nozzle and motor casing, and would provide much needed extra strength, resisting the immense pressure of the inside of the combustion chamber. Ideally, the smaller I can make the throat (narrowest part of the nozzle) the  the better. This is due to the fact that in order to generate the most thrust, a great amount of force is required, which is provided by having the exhaust gas acting over a smaller area. In order to decrease the the throat diameter, the nozzle must be able to withstand the higher pressure exerted on it by the rest of the exhaust gas which is trying to escape from the chamber. It follows that the strength of join between the nozzle and the casing (at least at my scale) correlates strongly to the maximum potential thrust that the nozzle can produce. Another problem that is evident with the plastic nozzles I have been using, is the fact that they rapidly deplete whilst burning - as in the plastic melts away and is expelled alongside the exhaust gas. This means that the narrowness of the throat does not always make a great deal of difference - if the throat is immediately blown apart before it has a chance to direct the exhaust gasses. 

With these two ideas in mind,  I proceeded to conduct a rather unscientific experiment. I wanted to test the effect of both the strength of the join, and the minimisation of plastic degradation. I got the idea for ablative protection from a documentary that I had watched on the space shuttle; I reasoned that by coating the nozzle in another material, it would provide a barrier between the plastic and the intense heat. I had already been doing this to an extent with every motor, by utilising a fine layer of bentonite clay, compacted between the propellant and the plastic nozzle, in an effort to prevent the immediate degradation of the joining. I then had the idea mix the clay with some water, in order to make a clay solution, which I could then apply to the nozzle, before allowing for it to dry - ideally leaving a solid layer of clay. To begin with, I submerged one of the nozzles entirely in a bath of dissolved clay - leaving the other one as a "control variable" despite the obvious fact that it has a different thread size - before removing it an putting it in the airing cupboard to dry. Unfortunately, this did not leave the desired amount of clay on the nozzle. Instead, I added more clay to the solution - making it thicker - and painted this on to the nozzle. I then allowed for this to dry and it had the desired impact of coating the nozzle completely.

I used these motors in the newly designed test stand. The first fire was placed onto a slippery garden chair, meaning that it flew backwards when the engine got up to maximum pressure in a pretty awesome display. Unfortunately, the pad flew into the wall of my house, cracking one of the supporting beams. Since this very fest test, the crack which was caused has continued to propagate, leading to the firing stand being sadly very damaged, a few weeks later. Further, due to the collision with the wall - paired with the thrust from the motor - the thread of the motor sheared off inside the motor mount. I then had to use a pair of scissors to save the component by digging out the remaining plastic. For the next test, I placed a brick on the predesigned section in order to prevent this issue. The pad then functioned much better, and it was obvious that the thrust created by the motor provided enough force to cause the motor mount to slid within the casing. This was a very positive demonstration.

This test did luckily gather some information about the benefit of adding a larger thread; the nozzle with the larger thread was found intact and attached the motor after the burn which was a positive. From this point onwards, I use more thread for this reason. I did decide whether to increase the dimension of the thread itself to increase traction under stress, however I decided that this would compromise the structural integrity of the end of the motor, as the walls are very thin - I couldn't make the nozzle much wider without running out of wall to drill my interior thread into.

It did seem that from this test, the clay proved useful, as the throat of the nozzle was in comparably good condition when I found it. Overall, the tests were effective, and demonstrated some that some serious thrust was produced; I still had a long way to go however.

For my future static fires, I decided to design and build a more sophisticated launch pad, in order to more accurately measure the thrust generated by the motors. Further, I decided that the test platform must be much more robust than the cardboard contraption that I had used for the previous launch test. This was one of the first large scale prints that I had completed, taking just over 6 hours to print - for this reason I opted for a partly modular design, which enabled the replacement of the section I deemed to be most drastically effected by the heat of the motor during combustion. I designed the stand to have a "Motor mount" section which slid into the main test stand.

A major issue I identified with the previous set up was the fact that the motor was purely 'sat' within the firing tube, with no real solid connection to the set up. For this design, I opted for a screw mechanism, with the same thread size as that of the inside of the nose cones. This meant that I could use motors interchangeably with the test stand for static fires, and with nose cones for flight, without altering the thread designs. 

The motor mount was on rails which then slid into the housing, enabling it to slide. This sliding initially was very difficult as the clearance which I had left was not sufficient and - due to the large surface in contact - much friction was present. I overcame this with a great deal of sanding - using one of my mums nail files - accompanied with some grease which I had from my days of custom keyboard building. I did intentionally leave a small clearance - my reasoning behind this was to ensure that when the motor mount slid, it didn't slide at an angle; if it were to slide upwards on an angle into the housing, it may get stuck, causing larger resistance. I believed that this issue would only be compounded due to the large mass of motor hanging off the end off of the end at a significant distance from the pivot, causing a large moment and pushing the opposite end of the mount upwards into the housing. In the end - after the sanding - the dimensions proved perfect, and I am still yet to have any significant observable issues with the joining between the motor mount and the housing.

I left a hole in the back of the housing, to allow for the future instalment of a strain gauge, which I then hoped to wire into my Arduino and my laptop in order to measure the thrust produced by the motor. As of the time of writing this report - a good few weeks after the initial production of this unit - I have still not implemented a strain gauge, however i have recently purchased one, with the hopes of slightly redesigning the housing and implementing the technology. I do believe that this was help me very much to determine the best design philosophies to adapt moving forwards. In the future I also plan to perform some more propellant testing in order to identify the optimum formula of the fuel I am using and to compare and contrast other possible fuels; this would very much rely on the implementation of strain gauge technology into the system.

The structure of the whole assembly is elevated off the ground and consists of a multitude of connected cross beams. I did not want to connect the beams horizontally at the base, as I believed that this would have caused instability if the pad were to be placed on uneven ground - instead I left the beams unconnected, enabling the stand to be placed on uneven terrain.  I designed a large counter weight which was connected at a large perpendicular distance away from the motor mount, in order to counteract the moment provided by the motor. In retrospect, the design did not really require the use of this counter weight, as instead, I would place a brick on the flat section which I had added onto the back for this very reason - I suppose that it did not harm the design however, and it did at least provide a way of aligning the brick so as to be perpendicular against the direction of the thrust. Further, I do believe this to have added a marginal amount of additional structural support as it was attached to both the top.

I was very confident that this launch would go as planned - hence the naming of it as "Shockwave"; I wanted a cool name for the first successful launch. How naïve of me.

The launch pad functioned well as usual, however, at this point I was beginning to wonder whether a longer launch pad would provide more support for the rocket so that it can get up to speed before leaving; I believed this to allow for the fins to provide improved aerodynamic support due to the increased exit velocity. For this test, I had increased the size of the thread to combat the nozzle ejection problem identified in the previous static fire - this worked. The nozzle remained attached to the motor casing, however the actual bell of the nozzle was melted/blown off.

As shown in the video, the motor and nose cone survived very well; I thought this might have been due to the freezing cold temperature -  it was at this point in the year that snow and frost was always lurking - which would have caused the plastic to cool rapidly after combustion, maintaining the structural integrity. 

I was however still baffled at the reason behind the thrust offset issue. From the results of this launch, I deemed that it must have been due to the throat of the nozzle being blown open early in flight, allowing for the unven release of exhuast gasses. 

In this launch I followed off from a lot of the design principals of the previous versions but with a few differences, most notably the addition of a plastic screw in nozzle and additional length of fuselage. The fuselage consisted of one piece of PLA matte which I 3d printed and screwed onto the top of the motor - in the same way as the previous nose cones. The nose cone atop the fuselage was rather blunted at the tip; a goal here was reusability and I feared that a point tip would incur more damages after repeated launches than this blunter design. Further, I wanted to reduce the risk - although I am not sure how effective this would have been - of damage to structures and people in the case of a direct hit. The main reason that I added the fuselage was to alter the ratio between the centre of mass, centre of lift, and centre of thrust. I had theorised that the problem with stabilisation in the previous iterations was stemming from the fact that as the propellant burned, the centre of mass shifted dramatically up the rocket, until it was almost in line with the nose cone - wholly due to the lack of "rocket" present above the motor. I supposed that this meant that the rocket would be very top heavy in comparison to the centre of thrust, making it unstable in flight. For this reason, I though it necessary to increase the length of the fuselage; this would mean that when the centre of mass was at its highest, it would still only be found half way up the rocket - much closer to the base of the rocket when the extra height is accounted for - enabling greater stability at speed. At this point in my journey, I had not done much in the way of research, relying significantly on my use of intuition, likely hindering my understanding of stability and aerodynamic concepts.

Another reason that I had identified for the dramatic spinning from the previous launches was the fact that the hole that I had bored was never accurately located in the middle of the clay nozzle - even if it were, the angle at which the drill bit entered the powder meant that by the time the hole would be finished, the angle would be so severe as to come in contact with one of the interior walls of the casing. This would clearly mean that certain sides of the nozzle were subject to significantly more stress than others, which, when coupled with the immense heat and pressure, would almost certainly cause a RUD (rapid unplanned disassembly). For launch three, I used a screw in nozzle in conjunction with a layer of compacted bentonite clay - this was to provide a limited amount of protection to the inner wall of the nozzle from the ludicrous temperatures reached during combustion. The aim of this nozzle was not only to help channel thrust and reduce engine blow out, but to also provide a template to which I could apply my drill bit in order to bore the central hole. This did work rather well, enabling holes to be drilled relatively accurately, especially when compared to previous attempts.

Unfortunately, this launch was rather ineffective, for reasons which I am still not entirely aware of.

Due to the discovered failures of the nose cone, this was an area which I worked a lot on for the next iteration of my rocket. I spent a long time in my cad software designing and printing various threads in order to attach the nose cone by a screwing mechanism. This failed many times; I could never quite seem to get the right diameter and pitch of the thread until I discovered the boolean tool. This meant that I could make the thread on the motor side - by means of sweeping a trapezium sketch along a helix pattern - before joining it with the nose cone with a boolean and using subtraction to remove the thread shape from the interior of the nose. I then realised that a reason that the threads never worked in the past was due to the lack of offsetting (clearance) between the faces of the threads. This meant that there was too much friction between the parts leading to the thread being too tight and not useable. With the use of the boolean tool, I could add an offset to the faces which increased the clearance.  On the right you can see the countless failed attempts at thread production that I endured. 

I used multiple camera angles this time, with the close up angle being in an effort to capture the lift off from the launch pad - to identify any problem with the launch sequence. As I mentioned in the previous launch, I used bricks to elevate the pad off the ground, in order to reduce the effect of flames rebounding off the surface, damaging the launch pad and the nozzle of the rocket. The biggest improvement made - other than the attachment of the nose cone - would be the size of the fins. Here, I increased the wingspan, in an effort to bring down the centre of mass and increase aerodynamic stability. This increased aerodynamic stability was counteracted however by a problem with an offset of thrust. This lead to the rocket veering violently of course, flying around my garden in violent spirals. Luckily, the nose cone improvement proved impactful, and the thread was intact during recovery.

The launch pad once again survived, I think in part to the elevation off the floor which meant that the hot exhaust fumes had limited time to bounce upwards and deteriorate the base of the silo. Due to the immense heat of the combustion, the PLA plastic deformed rapidly, loosing its structural rigidity, which may have been a factor which contributed to the lack of stable flight, however it is my belief that the majority of the deformation occurs after the rocket hits the ground as there is only a limited amount of time in the air.

Once more in an effort to combat the nozzle ejection problem, I played around with increasing the surface area between the clay and the inside wall of the motor. This time I added a thread which was cut into the inner wall, which seemed as though it would yield a great increase in traction over the previous version, however, when the rocket was found, once again there was no nose cone in sight - certainly a let down. 

In order to work on my launch platform, I implemented many new methods of stabilisation. I completely reworked the launch pad, with the most notable improvement being the addition of the large wide base legs in order to raise the motor of the ground. This then allows me to access the fuse easier; I still however support the pad by means of bricks. My thinking was that I didn't want flames rebounding off the ground and damaging the nozzle - a proposed problem which the raptor engines on starship would have faced during landing if not for the landing tower catching the booster. 

I then added fins to the motor casing in order to provide aerodynamic stability at speed. These fins then slotted into grooves which I made into the side of silo. These were in an effort to keep the rocket pointing directly upwards when launching. My thinking was that as the motor left the silo traveling straight upwards, it would be traveling at a speed at which the fins would be able to act in order to prevent instability. I can not say for certain whether the failure was on the behalf of the fins, however, as I think there are many other factors at play here. Despite this, I figured that larger fins could only help with the aerodynamic profile and bringing the centre of mass of the rocket downwards. Further, I drastically reduced the inner diameter of the launch pad resulting in a snugger fit . This was in an effort to retain a straight angle during launch and to reduce snagging on the walls of the pad. Further, the nozzle was not present at recovery, indicating that it was ejected at some point during flight, likely due to the lack of traction between the clay and the PLA which faltered under the high chamber pressure, combined with the high GeForce's experienced during the erratic flight. I had tried to score the inside of the motor using a fine pair of scissors in order to increase the surface area in which the clay and plastic would meet - hoping that this would help to eliminate the nozzle ejection issue which I observed on the previous test - however this proved futile.

With this first rocket, I simply hot glued the nose cone onto the top of the motor, however this proved to be highly ineffective. Firstly, the G Forces pulled by the rocket during launch - especially when launching at strange angles - easily overcame the strength of the glue joining. This issue is only exacerbated by the fact that the motor itself reaches incredibly high temperatures during combustion, resulting in the glue weakening during launch. Further, the glue is actually very weak for the amount of strength it provides, and it required a great thickness of nose cone in order to be effective - thus bringing the centre of mass upwards and fuelling the instability issue. I had made a miniature model of the rocket and launch pad in order to test the clearance, it did work as intended, however before I could get a video of it working, I broke the side panels of accidentally - easily done considering their immeasurable thinness.

This was very promising for a first launch of the revitalised space program in the new year. This was actually the first time I had ever got anything of the ground and gave me a lot of motivation to continue to work scientifically on the project. I got some useful information from this launch and really began to understand the importance of nozzles and chamber pressure. This also helped me to understand the importance of the launch pad - in particular the stability and support that it provides the rocket. I decided that if I were to get a rocket going straight up, I would have to seriously work on my launch pad. 

The reason this launch failed so catastrophically was  due to the way the motor got stuck in the silo, leading to a build up in pressure and then a launch at an angle. The reason the motor got stuck was due to the excessive clearance between the motor walls and the silo and the lack of support of the bottom of motor - this meant that when the motor launched at a slight angle, the bottom of the casing got stuck against the wall. Further, the nozzle ejected, marking the beginning of my ejection issues...

On the plus side, the flames out of the makeshift flame trench were pretty and the launch pad survived

This was my first attempt at designing a part for a practical purpose, and for a first attempt it was not the worst. I aimed for there to be a lip on the inside for the edge of the rocket motor to rest upon, so that I could then light a fuse from underneath. The exhaust gasses will then exit from the bottom of the motor and out through the  bottom of the launch silo. For these gasses to properly vent - and for me to be able to light the fuse from beneath - I had to then support the structure off of the ground with bricks.

This was my very first motor of my newly started rocket program and actually came about by chance; at this point I was still not sure whether 3D printing would be a viable means of construction for rocket parts - due to the high temperatures reached. This first tube was not intended to be a rocket at all - in fact it was solely for practice for me with my 3D printer. During a quick clean out of my cupboards - attempting to clear room for my filament spools - I discovered an old wooden percussion block which just so happened to fit perfectly within the interior diameter of the cylinder. I ran downstairs, grabbed my rocketry gear from the garage, and hurriedly assembled a motor. I used bentonite clay of an unrecorded depth as a nozzle and used a 3/16 inch drill bit to had drill a hole through the compacted clay powder and 35:65 sorbitol : potassium nitrate powdered fuel mixture. The motor was printed out of PLA plastic.