Project Blacksky 200K two stage - Class 3 submitted!

[Xrain]

Correct me if I'm wrong but it depends on the board's orientation. If it's measuring acceleration only in the Y direction, the acceleration before arcing over in relation to the board will be negative. However, once it arcs over the rocket will continue to have the same negative acceleration but because the board's orientation is reversed, the board will read it as positive. During that arc as the board's perceived acceleration changes from negative to positive. At some point it should hit 0 acceleration in the Y direction. Deployment can be triggered at that point. Of course this won't work if there is backsliding because the board's orientation will always be the same.

Alex

As soon as the motor burns out, and it is sufficiently far out of the thick part of the atmosphere the rocket will be in freefall. Meaning any gyroscope Accelerometer (oops) will be measuring 0 g's of acceleration for the entire period. Yes gravity is still acting upon the rocket of course, but as far as the internal reference frame of the rocket it will see no acceleration. It is very similar as to why you experience zero gravity on the vomit comet. So he is right you will not be able to get orientation from acceleration. You would need a gyroscope, horizon sensor, or perhaps magnetometer for that.

There is another issue as well, the rocket might not actually turn around until well into its return. So its very possible it could be going several mach before it actually turns around. It isn't the most probable case, but it can definitely happen. At 200k ft the rocket is in almost high vacuum. There are not going to be much aerodynamic forces until the rocket gets back to ~100k feet. I'd say the most likely case with out roll stabilization is the rocket will start to tumble on ascent at about 150k ft. If you don't want spin stabilization because it will make everyone sick watching your video, put 3-4 gopros on board, one in each direction. Then you can stitch portions of the frames together afterwards to eliminate the roll from the video.

Titanium can be had pretty cheaply from titaniumjoe.com I used a fair bit off it from there. Also surface grinding large thin titanium plates would be difficult since it won't hold to a magnetic chuck. I spend a lot of my time at work simplifying parts for low volume manufacturability and I agree that you'll want to simplify your parts to remove multiple machining setups, speed roughing, and avoid complex cnc programming. In small quantity runs these things will make an order of magnitude or more impact on costs.

I was taking a total shot in the dark on the surface grinding. I haven't actually personally done any surface grinding; but I know it has been used for similar kinds of cuts, holding it down will definitely be a challenge. I cant honestly think of that easy of method to make those large bevels on the fins without making a fairly complicated work holder.

I've looked at titaniumjoe's website quite a bit, but I haven't really had a need for much titanium yet. That will probably be changing in the not too distant future so I'll definitely check them out when I do finally bite the bullet.

 

[Aksrockets]

As soon as the motor burns out, and it is sufficiently far out of the thick part of the atmosphere the rocket will be in freefall. Meaning any gyroscope will be measuring 0 g's of acceleration for the entire period. Yes gravity is still acting upon the rocket of course, but as far as the internal reference frame of the rocket it will see no acceleration. It is very similar as to why you experience zero gravity on the vomit comet. So he is right you will not be able to get orientation from acceleration. You would need a gyroscope, horizon sensor, or perhaps magnetometer for that.

There is another issue as well, the rocket might not actually turn around until well into its return. So its very possible it could be going several mach before it actually turns around. It isn't the most probable case, but it can definitely happen. At 200k ft the rocket is in almost high vacuum. There are not going to be much aerodynamic forces until the rocket gets back to ~100k feet. I'd say the most likely case with out roll stabilization is the rocket will start to tumble on ascent at about 150k ft. If you don't want spin stabilization because it will make everyone sick watching your video, put 3-4 gopros on board, one in each direction. Then you can stitch portions of the frames together afterwards to eliminate the roll from the video.

Right you are, I guess I'm all physic'd out for the year. I'd imagine it could still be done with an accelerometer in the X axis, parallel to the horizon that could monitor any tilt in that direction.

Alex

 

[Frozenferrari]

Yes, this was my understanding for accelerometer based detection. Wouldnt the board still have no problem detecting when the rocket has "tipped over sideways" due to the value of X axis going from 0 (vertical during flight) to 1.0G when horizontal? Like turning a phone on its side. Help me understand what I am missing here...

 

[Xrain]

Xrain-
First off, I sincerely appreciate the in depth post! To start off, those numbers you see on my budget are pure guesses. They would be all inclusive, meaning the exact cost to go from drawings on my laptop to part in hand. I have semi-adequate models in .DWG format. Chris has graciously offered assistance in making more machinist-oriented models in more adaptable formats, as well as attempting to help me locate a shop willing and able to work with me. This being said, my current designs may indeed need a re-vamp to be more machine friendly.

My thinking is that for the Interstage, the majority of the work can be done first on a CNC lathe, and then 5 axis mill or wire EDM to cut the slots between the tongues and add the chamfers to them. However, I am not sure at all that this would be as easy as it sounds, or even possible. It is again just a guess on my part. The only machine experience Ive had before was with 3 axis milling.

I am stiff on changing the design of the interstage away from single piece machined, but I am not completely closed to the idea of a re-design. My thinking is that the tongues could maybe be engineered to slide into slots and then be locked in with a pin, or set screws, or possibly even bonded. But my worry with this design is weakened overall strength and integrity. What do you think?

I really do not want to change the weight or length around too drastically as everything must be redone for every single little change. All simulations must be re-run, and all models updated, to once again prove that the design will remain stable under all operating regimes. But more importantly, it seems that total length comes at a high price with this rocket.

Yes, the fincan is quite similar indeed to what you have done in the linked thread. I will for sure read that whole thread very closely and thoroughly tonight. Where did you get the 5" dia number from? The fin can sleeve can be machined from a 4" diameter rod or tube with .5" wall. Originally, I had the fillets to be 1/2" wide and 1/2" tall, but as Chris was explaining to me, a slight height reduction allowed for the stock to be an even 4" .

Indeed, another thing we had discussed was the ease of machining of the fins. As I am learning, a precision machine job putting on a true diamond airfoil and bringing them down to super thin tips would indeed be exceedingly hard with titanium, so I am looking at alternative shaping. I think I may just mill a standard hexagonal airfoil and allow the LE and TE thickness to be what it must to keep machining costs reasonable.

I should start again from square one with the interstage. A clean slate. Again, I REALLY want to keep the length the same, and the sleekness. But I am just not seeing how. Maybe it will come to me. I think making the IST two parts (tongues and transition) will dramatically help, but I dont see it fixing the issue entirely. Im open to any and all suggestions for changing it.

Again, I really appreciate everyone's time and thoughts they are putting into this with me. The rocket was founded on teamwork and will continue to evolve and live all the way to launch day on teamwork. Any input is good input. If you have ideas pitch them to me, I need em!

Man you like all the fancy expensive machining tools. That is fine, but all of those will cost a lot more to have your things made on them than other tools. Might as well go whole hog and do some friction stir welding as well.

Very simple EDM cuts start at around $500 just to cut a ~4" diameter part in half. There are probably cheaper places but not that much cheaper. Also very few machine shops are likely to have all that equipment. With exception of some very specialized machine shops, most are unlikely to have a EDM machine in conjunction with CNC machines. I would very strongly suggest makeing all of your parts possible to be made on manual machines. It will open up your possible machine shops considerably. It will probably help finding a CNC shop that is willing to make them as well, since if they only have to take one of their money making machines out of paying work for a day or so to help you, vs. a week they are much more likely to want to help you. A 5 axis would also be one of their flagship machines if they have one, since prices on them start at a quarter million. A shop like that might have a smaller 3 axis off to the side that they would be more willing to free up.

Welcome to rockets! Where every change on anything affects everything else! I understand you on this since it is quite a pain in the rear to redo all your sims and calcs, but I don't feel like you have quite fleshed out all the details on your rocket sufficiently to lock out design changes on it. If you had a portion of the parts already made that is one thing, but with just sims and CAD drawings now is a good time to really put all your design ideas through the ringer while it still doesn't cost anything (other than your time).

I say these things because I know you are trying to do a kickstarter, but I wouldn't put all your hopes in it considering the resurrected Armadillo Aerospace cant even get enough kickstarter funds to even match your budget. Have some backup plans so incase that one funding avenue doesn't work out, you still have a chance at accomplishing it.

 

[Xrain]

Yes, this was my understanding for accelerometer based detection. Wouldnt the board still have no problem detecting when the rocket has "tipped over sideways" due to the value of X axis going from 0 (vertical during flight) to 1.0G when horizontal? Like turning a phone on its side. Help me understand what I am missing here...

So I am probibly messing up the exact details on this so forgive me physicists. In the atmosphere your assumptions are reasonable. For one, it is reasonable to assume that your rocket is going to turn around when it hits apogee.

Outside of the atmosphere (You will be outside 99.9% of it) It isn't even reasonable to assume you will be sideways at apogee. Since you have for all intents and purposes absolutely zero aerodynamic forces on your rocket. You cannot count on anything that your fins will do for you.

This is a playback of a telemetery stream on a sounding rocket that my university launched. It went to a ~97 km apogee. The video is displaying all its position and attitude information. As well as the heat levels on the nosecone. I think the data output on the left of the screen is from a plasma probe if I remember right (honestly cant really remember what it is).

[YOUTUBE]OZjPvpcWt8I[/YOUTUBE]

If you notice on that video the rockets attitude does not go horizontal at apogee. There are a couple of reasons for this, main one being it is spin stabalized so it has some natural stabalizing force. Two there are no real forces acting upon it.

The rocket doesn't actually turn around until it hits the stratosphere again, doing close to mach 3. Your rocket will be going much slower at the high altittudes so you will not have even close to these forces acting upon it.

Its turn around is shown in this video.

[YOUTUBE]SjV8fdqVZ2g[/YOUTUBE]


So to even start with, your base assumption that the rocket will be horizontal at apogee isn't correct, without any kind of intentional stabilization like spin, or active control. You cannot assume that it will be in any particular attitude.

Not to mention that being in free fall it wont actually have any acceleration forces that the rocket itself can measure. At apogee for any rocket if I am not mistaking something is at 0 acceleration since all the forces acting upon it negate each other.

Also most phones use a 3 axis gyroscope and often a 3 axis magnetometer in addition to the 3 axis accelerometer to get its attitude information.

 

[Zebedee]

On a different topic - A "normal" kickstarter project is something where people invest in order to get a "product" at the end. What "product" are you offering folk who invest in this project?

 

[JimJarvis50]

Good luck on the project Ryan. Very ambitious I would say!

I see you are using a T charge. Good application I think. However, I don't think you have enough room for it as you have drawn it up. The device works because there is some backpressure generated from the legs. Cut the legs too short and it won't work. You have it located in the constricting end of a 3-inch cone, and I don't see how there is enough room. However, you should be able to use the 1-sided approach I use now. Maybe it wasn't in the article you read (I have made additions). The most recent is attached, with the section of interest at the end.

Jim

View attachment Article on high altitude deployment charges_May 2013.pdf

 

[prophecy]

Easy to get lost in the minutiae, but taking a step back bigger picture - it's a 4"-3" flying commercial motors (which can only be optimized so much because the design is geared towards reusability and ease of use for the consumer as opposed to optimization), with a significant amount of metal parts elsewhere, shooting for 200k?

Jim Jarvis and Neil Anderson spent years and tens of thousands of their own money perfecting the staging commercial motors concept that you're hoping to step in and nail on the first go, in three months, with other people's money - and double their altitude. What gives?

Sounds like you've spent a lot of time thinking it all out - good for you - but if conceptual and actual were the same thing, my Black Rock project is simming to 240k - I know that I'll be lucky to hit half that. And that's with a case-bonded motor in a carbon case and a mass fraction well beyond what is doable with commercial motor setups.

Wish you the best of luck and I look forward to seeing the attempt. If you get this to the pad, good for you - the means are as ambitious as the claim (building a pimped out project with very little actual funding in a three month span). If you not only get to the pad, but have a perfect flight - another big if - you'll be in the bottom half of the 100k range.

Have fun.

-steve

 

[bobkrech]

Xrain

That's exactly what GoFast did in 2004 according to their presentation at NARCON 2009. It was IIRC spin stabilized and went straight up to ~ 370kft and then backslide down ~ 150 kft until the lower atmosphere flipped it over.

Bob

 

[Frozenferrari]

Steve - I shared your doubts of the '200K' number at first too. Especially with fast burn motors. It was the Loki Dart system that convinced me of the possibility. A P-class motor expended below 10,000 feet resulted in a low drag, high inertia dart going 270,000. Lower drag values help. AeroPac100K was able to send a 4" to 3" rocket to 104K with 21,000Ns. Im using almost 50% more, in a much shorter and lighter airframe. Do I believe the 250 something forecasts out of OR? or even the 230s from RAS? No. But I am reasonably confident in 180-220K . So I leave my projection where it stands at 200,000 feet, with a waiver to 300K just in case.

Bob and Xrain - You bring up a fantastic point I hadnt considered, backsliding. I was counting on the rocket to tip horizontal at apogee such as we saw with all the 100K flights, which would allow the accelerometer to register an X value and therefore actuate deployment. Thank you Xrain for that video, that really drove it home for me actually. Okay, this is a significant issue but I have a feeling we can work around it with some careful thinking.
I really do not want to install gyroscopic stabilization systems due to their space and weight requirements, but I definitely should atleast give it some fair research. I will do some Googling on these systems, beginning with what are commonly used on such rockets as the SpaceLoft systems. Are there more readings that come to mind I should be aware of?
So let me ask this - What would induce the tumbling in the first place? As we saw with some of the 100K flights, they remained pretty uniform in motion all the way until they arc'd over at apogee. But none of these passed 150K, which is where you had suggested I might get tumbling. * I am assuming that this is because the small amount of remaining atmosphere between 60 and 120K (where most of those birds apogee'd) is enough to dampen out non vertical flight, and induce some correctional moment, which of course carries beyond the targeted point and then results in a new correctional force coming from the other side (positive static and negative dynamic stability?), but when you get past 150 there is not enough and the rocket just keeps swinging farther and farther out of vertical each time untill it is tumbling. Is this accurate, or even on the right track of thinking?
Here's what comes to mind for working around this WITHOUT the addition of gyroscopic systems-
I cannot be sure whether the rocket would indeed start tumbling, or if it would stay vertical and then backslide, so therefore I cant count on an accelerometer turning sideways and picking up that X value. So this leads to the next best form of apogee deployment logic, a timer. But the problem with a timer is that it cannot be relied on as the only logic. If I fail to light the sustainer motor, a long timer would allow the vehicle to come in hot.
Is there a way to program any existing altimeter systems (or better yet, the Raven, the Telemega, and the Aim Xtra) to watch for sustainer ignition G, and then set a timer based on this, OR if the motor does not light, set a much shorter timer and/or watch accelerometer data and fire at apogee still?
Another thing that comes to mind is inertial reference systems. I am aware that the Aim Xtra can plug in inertial data to validate it's GPS signals, so perhaps the system could mathematically find apogee based on detected G from flight? Even if this werent super accurate, it could work, I think.
I could always choose to deploy early on purpose, as the lack of drag above 100,000 should allow for this without damaging the rocket. But it just seems "done-wrong" to purposefully allow the sustainer to deploy well before apogee. Furthermore this could have negative effects on the video, yanking and jerking it around.

About the machine work. I learned last night of a relatively new sort of metal production process, geared toward speedy and cheap production of complex parts. It's called DMLS, or Direct Metal Laser Sintering, and from what I can tell, it's sortof like 3D printing a metal part. I want to do more research on it this weekend. I want to try and contact a few shops that have the ability to do this work, and perhaps see if my interstage transition (maybe even the fincan too!) could be produced with this process for a reasonable cost.

Keep the thoughts coming guys, I really appreciate the feedback. I think at this point it might be best to suspend my launch date. I knew it would be tight as is, and I expected this possibility, but I can tell I may have more work left to do than I thought. AeroNaut, XPRS, and BALLS are all on the table for launch events this year, as they all have great launching weather in the latter half of the summer when the air is nice and warm, IF the rocket flies this year afterall. This is still my goal if I can help it.

 

[Frozenferrari]

Good luck on the project Ryan. Very ambitious I would say!

I see you are using a T charge. Good application I think. However, I don't think you have enough room for it as you have drawn it up. The device works because there is some backpressure generated from the legs. Cut the legs too short and it won't work. You have it located in the constricting end of a 3-inch cone, and I don't see how there is enough room. However, you should be able to use the 1-sided approach I use now. Maybe it wasn't in the article you read (I have made additions). The most recent is attached, with the section of interest at the end.

Jim

I had seen it.
I chose not to employ the single direction method because it A- protrudes into the recovery bay a little more, but this can be worked around. I suppose I could even design it so that it is recessed entirely into the avbay and doesnt protrude at all. Hmmmm. Now youve got me thinking!!
and B- because I did not want to turn any of my recovery or my tracker into a bullet. I learned with my 54mm rocket that an undirected charge can propel things fast enough in the nosecone to break tracker housings and avbay housings. I was thinking that the T method would direct gasses perpendicular and away from these things. What are your thoughts on this?

Are you sure youre looking at the right image? I have the tee mounted to the top of the avbay lid, which is near the base of the nosecone. It is pretty wide down there. Though I will go back, remeasure, test, and remodel to be sure.

 

[Xrain]

Steve - I shared your doubts of the '200K' number at first too. Especially with fast burn motors. It was the Loki Dart system that convinced me of the possibility. A P-class motor expended below 10,000 feet resulted in a low drag, high inertia dart going 270,000. Lower drag values help. AeroPac100K was able to send a 4" to 3" rocket to 104K with 21,000Ns. Im using almost 50% more, in a much shorter and lighter airframe. Do I believe the 250 something forecasts out of OR? or even the 230s from RAS? No. But I am reasonably confident in 180-220K . So I leave my projection where it stands at 200,000 feet, with a waiver to 300K just in case.

Bob and Xrain - You bring up a fantastic point I hadnt considered, backsliding. I was counting on the rocket to tip horizontal at apogee such as we saw with all the 100K flights, which would allow the accelerometer to register an X value and therefore actuate deployment. Thank you Xrain for that video, that really drove it home for me actually. Okay, this is a significant issue but I have a feeling we can work around it with some careful thinking.
I really do not want to install gyroscopic stabilization systems due to their space and weight requirements, but I definitely should atleast give it some fair research. I will do some Googling on these systems, beginning with what are commonly used on such rockets as the SpaceLoft systems. Are there more readings that come to mind I should be aware of?
So let me ask this - What would induce the tumbling in the first place? As we saw with some of the 100K flights, they remained pretty uniform in motion all the way until they arc'd over at apogee. But none of these passed 150K, which is where you had suggested I might get tumbling. * I am assuming that this is because the small amount of remaining atmosphere between 60 and 120K (where most of those birds apogee'd) is enough to dampen out non vertical flight, and induce some correctional moment, which of course carries beyond the targeted point and then results in a new correctional force coming from the other side (positive static and negative dynamic stability?), but when you get past 150 there is not enough and the rocket just keeps swinging farther and farther out of vertical each time untill it is tumbling. Is this accurate, or even on the right track of thinking?
Here's what comes to mind for working around this WITHOUT the addition of gyroscopic systems-
I cannot be sure whether the rocket would indeed start tumbling, or if it would stay vertical and then backslide, so therefore I cant count on an accelerometer turning sideways and picking up that X value. So this leads to the next best form of apogee deployment logic, a timer. But the problem with a timer is that it cannot be relied on as the only logic. If I fail to light the sustainer motor, a long timer would allow the vehicle to come in hot.
Is there a way to program any existing altimeter systems (or better yet, the Raven, the Telemega, and the Aim Xtra) to watch for sustainer ignition G, and then set a timer based on this, OR if the motor does not light, set a much shorter timer and/or watch accelerometer data and fire at apogee still?
Another thing that comes to mind is inertial reference systems. I am aware that the Aim Xtra can plug in inertial data to validate it's GPS signals, so perhaps the system could mathematically find apogee based on detected G from flight? Even if this werent super accurate, it could work, I think.
I could always choose to deploy early on purpose, as the lack of drag above 100,000 should allow for this without damaging the rocket. But it just seems "done-wrong" to purposefully allow the sustainer to deploy well before apogee. Furthermore this could have negative effects on the video, yanking and jerking it around.

About the machine work. I learned last night of a relatively new sort of metal production process, geared toward speedy and cheap production of complex parts. It's called DMLS, or Direct Metal Laser Sintering, and from what I can tell, it's sortof like 3D printing a metal part. I want to do more research on it this weekend. I want to try and contact a few shops that have the ability to do this work, and perhaps see if my interstage transition (maybe even the fincan too!) could be produced with this process for a reasonable cost.

Keep the thoughts coming guys, I really appreciate the feedback. I think at this point it might be best to suspend my launch date. I knew it would be tight as is, and I expected this possibility, but I can tell I may have more work left to do than I thought. AeroNaut, XPRS, and BALLS are all on the table for launch events this year, as they all have great launching weather in the latter half of the summer when the air is nice and warm, IF the rocket flies this year afterall. This is still my goal if I can help it.

I've been involved with DMLS fabricated parts at my work in the form of 3d printed hot gas thrusters, as well as I got to see SEDS 3d printed liquid engine and help them a little bit in firing it. DMLS is fast, but it currently isn't cheap. A 10-12 in long part was ~$20,000. It can definitely work to make good parts, the surface finish however might require post work on your mating faces for your interstage since it isn't all that smooth. The high price is from the fact that the printers run in the 1-3 million+ range to purchase one. So they are really expensive to get time on.

So when I say gyroscope I do not mean a stabilization system, I think you are thinking of reaction wheels. The gyroscopes I am talking about are a sensor that is used to measure rotation. You can get them small enough that you could fit several on a dime.

Honestly your most reliable methods of detecting apogee would be an unlocked GPS, but these do cost a significant amount. The most reliable way I could think of to deploy a parachute would be a timer, and then fly an IMU along with it so you can collect flight data, to know how your sensors are going to behave. Then if you ever do another flight you will have some flight data that you can base you deployment logic on in the future. I personally just wouldn't have the confidence in my deployment system logic when there just flat out aren't enough flights to these altitudes. If it were me I would actually time it to deploy after apogee, since the longer the parachute is deployed the longer it has to tangle itself. You also have to remember that the higher you go the harder it is going to hit the thick part of the atmosphere on the return, so you need to design your parachutes appropriately. On flights to ~100k feet it doesn't pick up enough speed to notice the rapid increase in air density so it smoothly slows down. On flights to higher altitudes the rocket picks up considerable speed before encountering any significant atmosphere, so it gets some significant forces. You cannot think of this rocket in the same ways you think of one going to 20k feet. You are in a completely different realm of flight.

The tumbling I speak of is from small perturbations on the rocket that can come from pretty much anything. These are normally easily dampened out, but with the lack of correcting forces, its not unreasonable to expect your rockets attitude to vary randomly after ~125k feet.

If you are worried about failed sustainer ignition you can use a standard barometric based deployment computer, as well as a timer based one. Use some logic that will disable one or the other depending on a what the sustainer does. There are more elegant solutions but this is one option.

 

[JimJarvis50]

I had seen it.
I chose not to employ the single direction method because it A- protrudes into the recovery bay a little more, but this can be worked around. I suppose I could even design it so that it is recessed entirely into the avbay and doesnt protrude at all. Hmmmm. Now youve got me thinking!!
and B- because I did not want to turn any of my recovery or my tracker into a bullet. I learned with my 54mm rocket that an undirected charge can propel things fast enough in the nosecone to break tracker housings and avbay housings. I was thinking that the T method would direct gasses perpendicular and away from these things. What are your thoughts on this?

Are you sure youre looking at the right image? I have the tee mounted to the top of the avbay lid, which is near the base of the nosecone. It is pretty wide down there. Though I will go back, remeasure, test, and remodel to be sure.

View attachment 263888

Well, I confess I don't know exactly what you're trying to do with your recovery system. I assumed that the "T" was that brass thing in the picture and that you are trying to pressurize the space at the top of the cone. If that's correct, then just mount a single-sided charge facing upwards. Sorry if I'm not interpreting this correctly.

Hey, you might be interested in another thread of mine. It's about stabilizing high-altitude, multi-stage flights for the purpose of getting approval to do them.

https://www.rocketryforum.com/showthread.php?122042-I-could-use-just-a-little-guidance

Jim

 

[prophecy]

The professional motors you're referencing ran steel cases at pressures exponentially higher than what an aluminum cased CTI motor will run. Apples and oranges. Spend all the time you want getting your own waiver, knowing full well that three people (Jarvis, Anderson and Biba) who nailed similar flights got in the low 100s. If you execute everything flawlessly the first time (which none of them did, BTW) and get a similar result well below your 200k claim, prepare to be disappointed with a result you ought to be thrilled with due to talking the talk before walking the walk.

Have fun and see ya on the playa-

Steve

 

[Clusterphiliac]

Regarding deployment: what's wrong with just integrating the accelerometer data to get the approximate apogee? As other users have mentioned, at that altitude the aerodynamic forces on the rocket are nearly zero, so if the deployment is slightly early or late it shouldn't put excessive stress on any part of the vehicle, right?

 

[MClark]

If the chute is deployed very high the load will be low but as it falls the velocity will be very high.
With a drop from 200k it will be well over Mach as it goes by 100k. Many designs will collapse over M1 and then reinflate rapidly and repeatedly destroying the chute. Shock inflation.
Add to this it is very cold and the chute material will likely be frozen and brittle.
Do some research on materials and parachute designs for high speed.

M

 

[mikec]

I was counting on the rocket to tip horizontal at apogee such as we saw with all the 100K flights, which would allow the accelerometer to register an X value and therefore actuate deployment.

You're still not getting it. Accelerometers alone in free fall cannot sense orientation, period. You need to use integrated accelerometer data to predict apogee (Telemega can do it), or just use a timer.

 

[bobkrech]

What most hobbyists fail to realize is that rocket flights to >100+ Kft have been done for the last 70 years with sounding rockets. There are literally 10's of thousands of technical reports on these flight and you can easily access them via the internet and if you read a few of them you will see there's no reason to reinvent the wheel.

You also need to be aware that ALL unguided sounding rocket launched by DoD and NASA at any US government or international test ranges MUST employ spin stabilization to minimize dispersion and are launched at least 5 degrees off vertical to unambiguously define the trajectory and impact splash zones. The Loki Dart you mentioned spins at 20 RPS and that's why it goes up straight and doesn't tumble near apogee. Larger diameter rocket have lower spin rates but even the 4.5" Arcas is spun at 20 RPS to it does not tumble during it's flight. That is how the attitude and direction of an unguided rocket is controlled in flight.

Parachutes are not deployed above 100 Kft because they won't work there as the air is too thin, and they would simply wrap around the rocket on the way down. At 100 kft and velocities near Mach 1, there is enough dynamic pressure to inflate a parachute however you need specially designed drogues that have only one shroud line or you risk wrapping the chute around the rocket. Inverted trigonal or square pyramid parachute designs are frequently used to stabilize objects dropped from these altitudes until the main deployment sequence is initiated at 10 kft AGL or lower.

Virtually every smart phone made in the last 5 years contains a 9 DOF inertial platform (3-axis accelerometers, magnetometers, and gyros) coupled to a GPS assisted navigation system so the phone knows where and at what attitude it is at all times and knows where you want to go and how long it will take you to get there. My company makes a sub-pound quadrotor reconnaissance drone using the same smart phone components in it's inertial navigation system and a dozen or so in domestic commercial use are registered with the FAA as experimental multiengine aircraft. We use the same barometric pressure sensor used in the StratoLogger and the standard deviation of our altimeter is +/- ~10' from sea level to 20 kft when tested to the FAA altimeter standards. The GPS location is within ~1' of "gold standard" systems that cost hundreds of times more. The 2 rocket altimeter/tracking systems I previously mentioned employ similar technology, so there really is nothing to prevent you or anyone else from knowing where your rocket is at anytime and that makes recovery simple so you again don't have to reinvent the wheel, or homebrew an inferior system.

Bob

 

[Frozenferrari]

You're still not getting it. Accelerometers alone in free fall cannot sense orientation, period. You need to use integrated accelerometer data to predict apogee (Telemega can do it), or just use a timer.

Now I see what youre saying! Okay, so help me understand this then- As soon as the motor lights and starts building above 1.0 positive G, flight computers start running calculations plugging that G in to create a velocity prediction, and when the motor burns out, it experiences 0G minus the added drag, so somehwere in the negative G range. So it then begins subtracting 9.8m/s every second as well as factoring in drag detected in that additional negative G, to create a new velocity prediction in real time, continuously. And therefore apogee would be the point at which said calculated velocity goes from something positive to zero and then into the negative, correct?

In that case, wouldn't I still be able to reliably use altimeters to calculate, or shall we say predict, apogee pretty reliably? Bob, is this what you mean by not having to re-invent the wheel? I am indeed looking through NTRS at other sounding rocket reports again, trying to specifically see what they have to offer on trajectory.

On the concept of backsliding versus tumbling up to apogee and then down- What exactly initiates a gravity turn in the first place? I had always thought it was the result of some slight aerodynamic disturbance, which does not get dampened out as velocity decreases and fin's correctional authority becomes degraded. Am I wrong? Why is it that when we watch the video of SpaceLoft-8 , which used spin stabilization to around 220,000ft and then de-spun, it has a verrrry gradual turn, and makes a nice clean arc over at apogee, similar to what we've seen with Jim's FCY and the other 100K flights, rather than backsliding or tumbling? What's to say my rocket will do anything different? And why is it that the video that Xrain was showing us clearly illustrated that some rockets have no arc whatsoever, and do indeed backslide. Is this due to the resistance to changing attitude due to how fast it was spinning? I guess what Im really trying to understand is why my rocket would initiate a tumble in the first place.... Or really just, what exactly might happen to a body such as a rocket traveling upward with no aerodynamics in play, no spin, and no spinning instruments inside of it resisting changes in motion.

If the chute is deployed very high the load will be low but as it falls the velocity will be very high.
With a drop from 200k it will be well over Mach as it goes by 100k. Many designs will collapse over M1 and then reinflate rapidly and repeatedly destroying the chute. Shock inflation.
Add to this it is very cold and the chute material will likely be frozen and brittle.
Do some research on materials and parachute designs for high speed.

M

Yeah, that's a good idea for sure. I had done a tad of looking around at such chutes like RocketMan's supersonic drogues, but youre right, I need to go do some document reading and find out more about how it is done professionally. Ill do some more searching this weekend. One thing that immediately comes to mind is the employment of a drag device other than a traditional 'parachute' . You know reflectors that professional photographers use? With a springy frame, they can be folded and zipped up for storage, but they kindof pop open when you unzip them. Perhaps I could employ something like this, in a dome or disk shape. Something that has a bit of assistance staying open, that way there wouldnt be such harsh shock on the material. Just an idea, I like creative work arounds. Ill do some reading this weekend.

 

[mikec]

In that case, wouldn't I still be able to reliably use altimeters to calculate, or shall we say predict, apogee pretty reliably?

Certainly, no one has ever said you can't. Many accelerometer-based altimeters (such as the Telemega) can do this. Or you can just simulate the time to apogee and use a pure timer as Jim Jarvis has done. Since there is an effective vacuum at these altitudes, being early or late by tens of seconds doesn't matter very much.

 

[Frozenferrari]

Ahhh! Cool. So people were just trying to tell me that I was incorrect about physically how the accel-based altimeters FIND apogee? That it is CALCULATED rather than sensed or detected. Cool! Okay, well I feel more comfortable basing the logic on something that the altimeter can evaluate in real life, such as G forces, rather than simulations. Heres why: As Steve was pointing out- This thing may not do quite what the simulations forecast. Time to apogee could be greatly different than what we are predicting. A 130Kft apogee would place the rocket coming in hot if I used the same two minute apogee deployment delay timer that I would for a 200K flight.
Pure timer could serve as a decent backup though, yes. Indeed, deploying early will not hurt the rocket BUT it could cause problems that COULD hurt the rocket. The main issue that comes to mind is the tangling of the shock cords. Secondly, the longer it's apart, the more yanking and tugging it does. I think Ive designed my attachment methods to be able to take a hell of a yanking, but let us not get careless if we can help it. Thirdly, I need to find some type of assisted opening chute or drag device so that if i DO deploy early, or even properly, the drag device wont experience the shock of being snapped open and closed repeatedly on decent, as was mentioned.
Some thoughts are coming to mind regarding my ablative as well. Has anyone else used Rx2390? The system 'chars' at high temperatures to provide thermal insulation to the substrate below. Now, this 'charring' sounds like it may involve carbon. If my nosecone has a layer of carbon formed over it from the heat, this may suddenly block GPS and RF signals. I would wonder if this 'char' is indeed conductive, and if it would do this as I am forecasting.... Regardless, this shouldnt be a problem so bad as to run out and make a change though. Yes, we will loose communication with the systems on the ground, but they should come back after apogee once the nosecone comes off the top of the avbay (however it might still be degraded as the LOWER half of the avbay is permanently surrounded in airframe which would have ablative on it) correct? Thoughts?

On a different topic - A "normal" kickstarter project is something where people invest in order to get a "product" at the end. What "product" are you offering folk who invest in this project?

Yes, that brings up a good point- I need to decide on a funding avenue and exactly what I can do for those who fund me. I could probably pour in a couple hundred to a grand or so of my own, but I will still need what, 12 more? Right off the bat I can tell you that those who donate will receive full and unedited versions of the onboard flight video, perhaps a cool metallic printed image of the liftoff.. or some exclusive directors-cut style footage taken during/at the launch. Another thing that came to mind is Tshirts. I kindof like the idea of making a small batch of them for the launch, and Ive heard a few people say they want one already. I feel like a team Tshirt would make a nice addition to the thank you packages for larger contributors.
But what will you all really GET out of it? Lets be honest. Just a cool video and to help us expand the ARLISS Xtreme program.... and maybe a Tshirt, DVD and a metallic image. Thoughts?

 

[ksaves2]

Question Bob: Does it follow then that at these extremes it's best to consider recovery of the sustainer and payload bay separately like CSXT as opposed to restraining the components via shockcord like use "lower flying" masses? Kurt

What most hobbyists fail to realize is that rocket flights to >100+ Kft have been done for the last 70 years with sounding rockets. There are literally 10's of thousands of technical reports on these flight and you can easily access them via the internet and if you read a few of them you will see there's no reason to reinvent the wheel.

You also need to be aware that ALL unguided sounding rocket launched by DoD and NASA at any US government or international test ranges MUST employ spin stabilization to minimize dispersion and are launched at least 5 degrees off vertical to unambiguously define the trajectory and impact splash zones. The Loki Dart you mentioned spins at 20 RPS and that's why it goes up straight and doesn't tumble near apogee. Larger diameter rocket have lower spin rates but even the 4.5" Arcas is spun at 20 RPS to it does not tumble during it's flight. That is how the attitude and direction of an unguided rocket is controlled in flight.

Parachutes are not deployed above 100 Kft because they won't work there as the air is too thin, and they would simply wrap around the rocket on the way down. At 100 kft and velocities near Mach 1, there is enough dynamic pressure to inflate a parachute however you need specially designed drogues that have only one shroud line or you risk wrapping the chute around the rocket. Inverted trigonal or square pyramid parachute designs are frequently used to stabilize objects dropped from these altitudes until the main deployment sequence is initiated at 10 kft AGL or lower.

Virtually every smart phone made in the last 5 years contains a 9 DOF inertial platform (3-axis accelerometers, magnetometers, and gyros) coupled to a GPS assisted navigation system so the phone knows where and at what attitude it is at all times and knows where you want to go and how long it will take you to get there. My company makes a sub-pound quadrotor reconnaissance drone using the same smart phone components in it's inertial navigation system and a dozen or so in domestic commercial use are registered with the FAA as experimental multiengine aircraft. We use the same barometric pressure sensor used in the StratoLogger and the standard deviation of our altimeter is +/- ~10' from sea level to 20 kft when tested to the FAA altimeter standards. The GPS location is within ~1' of "gold standard" systems that cost hundreds of times more. The 2 rocket altimeter/tracking systems I previously mentioned employ similar technology, so there really is nothing to prevent you or anyone else from knowing where your rocket is at anytime and that makes recovery simple so you again don't have to reinvent the wheel, or homebrew an inferior system.

Bob

 

[Frozenferrari]

Question Bob: Does it follow then that at these extremes it's best to consider recovery of the sustainer and payload bay separately like CSXT as opposed to restraining the components via shockcord like use "lower flying" masses? Kurt

Not going to happen. Would require 2 Avionics bays, 2 recovery bays , 4 GPS systems, 2 RF systems and it overcomplicates recovery processes. Rocket would no longer go as high as it would be much larger and heavier.
I do still want to hear Bob's thoughts on this though. I think the tangling could become a big concern indeed. Perhaps it would help to add some sort of tangle resistant sheathing to the cord, but that might make it harder to pack.... One thing I can guarantee is that the main parachute will need to be placed VERY close to the avbay shock cord connection so even if it does turn into a bird's nest, there's still a reasonable chance that the main parachute will open and do it's job properly. Also, Shock cord is more conforming than a balled up parachute, so Id rather stuff that into the nose tip and reserve the wide part for the chutes and the CD3.
Another thing to consider is Ematch wires breaking, or being ripped out of position during the tumbling. On previous rockets I had found that occasionally tying a knot with the Ematch wires and the shock cord (with some slack left in the match wire) prevented the wire from ever becoming tensioned. Shrink wrapping or taping it, or feeding it through the cord, will also help keep it from tangling as well as protect it from bumping up against anything.

I have a friend helping me recreate models of the Interstage transition in STEP, as all I have is AutoCAD and it cannot export as a STEP. From there I want to run it through the instant quote you can find here: https://www.nextlinemfg.com/metal-laser-sintering/?gclid=CKjbrOiy1cUCFYVFaQodPXUAoQ
Assuming Xrain is right, and it will cost an exorbitant amount, I will then redesign. But if by some miracle I can find a shop that will print it for under $3000 or so, Id be interested. I have a feeling when the quote comes back its going to be on the higher side of that number, especially given it has been stated from experience that it might be. But lets say it isnt-- What sort of surface finishing you mentioned might I actually be looking at to prepare it to mate with the other parts nicely?

 

[A5tr0 An0n]

I have a friend helping me recreate models of the Interstage transition in STEP, as all I have is AutoCAD and it cannot export as a STEP. From there I want to run it through the instant quote you can find here: https://www.nextlinemfg.com/metal-laser-sintering/?gclid=CKjbrOiy1cUCFYVFaQodPXUAoQ
Assuming Xrain is right, and it will cost an exorbitant amount, I will then redesign. But if by some miracle I can find a shop that will print it for under $3000 or so, Id be interested. I have a feeling when the quote comes back its going to be on the higher side of that number, especially given it has been stated from experience that it might be. But lets say it isnt-- What sort of surface finishing you mentioned might I actually be looking at to prepare it to mate with the other parts nicely?

You referenced SW in your Microsoft Word write up, so if you have access to it then you can convert your DWG file to a STEP file format within SW. Pretty simple and easy to do. Also DMLS parts typically have a raw type finish as previously mentioned. Most are about the equivalent to a medium turned finish or 350RA microinches. So you would want to have the part with a finer surface condition to allow a smoother fit and reduce issues that could result with the increased friction. But I am no expert with DMLS.

As far as this project goes I see nothing wrong with making big jumps and advancements as you are doing to further your rocketry state and the state of others. With that said I do see some issues with some of the design mainly on a construction and economic stand point. In your MS word write up you also mentioned running computational fluid dynamics, have you ran any (specially any equation based goals)? Seeing as all your simulations which are in RAS and OR, which only allow a "standard," transition, your Cd will be off and can be significantly off. Why? Because OR makes the following assumptions, boundary and equation conditions are set for large areas/volumes, base drag is corrected by subtracting the motor base area (due to motor exhaust disrupting vortices formed in that region), and the boundary layer assumption is turbulent throughout. The reason I quote that is because I am curious if you have come up with accurate external values for you Cd and imputed them into OR/RAS? I am skeptical that the increase in altitude (assumption) is actually worth the added cost in manufacturing, especially since you lack the ability to ,machine parts in house (costs will rise).

Seeing as your reasoning (based on your MS Word write up) for utilizing a two stage configuration (albeit more efficient) was that it is cheaper than a single stage configuration. With you forecasting ~$13k for the total project, that is no longer the case. A single stage vehicle to the same altitude can be made for far less than that. If you were putting more than ~1/13th the cost up front then I wouldn't think that the financials mattered as you can financially swing it. However that is not clearly the case. I am not trying to slam you or disrespect you or anything negative for that matter, just merely stating that I think you will need to go back to the drawing board and redesign some parts while rethinking the choice of some materials; if you ever want to see this project get off the ground or out of the computer. I am skeptical that you will get even close to enough funding (via kickstarter) to support your current goals and no way is this coming together in time for Aeronaut at least in its current design. If so then all power to you man, you rocked it and I look forward to seeing you there! Just thought I would add my opinions.


Cheers!

 

[Frozenferrari]

You referenced SW in your Microsoft Word write up, so if you have access to it then you can convert your DWG file to a STEP file format within SW. Pretty simple and easy to do. Also DMLS parts typically have a raw type finish as previously mentioned. Most are about the equivalent to a medium turned finish or 350RA microinches. So you would want to have the part with a finer surface condition to allow a smoother fit and reduce issues that could result with the increased friction. But I am no expert with DMLS.

Indeed! One of my colleagues who was assisting me design the rocket was making models and running CFD in SolidWorks, but when he left the team I lost access to the software. I guess he lost his drive, and I agree it is an arduous process, but like it's been said.. Thats rocketry. The only CFD render I ever actually was able to secure is one showing dynamic pressure on the nosecone at high mach to justify placement of barometric relief ports and the viewing window for the GoPro. Cd values for the full stack could be obtained from RAS running a full stack profile, but lacking sustainer fins. I wish I had solidworks. All I have is AutoCAD. This being said, I am having my friend Chris C. help make updated models and perhaps he may be able to run some stress analy on the tongues and CFD on the stack, and I may finally get those renders. Depending on the resuslts, I may be able to run some lesser expensive materials for the fincan and IST, such as 6061 or something else. I have a strong feeling the fins will need to remain titanium, as they go far beyond forecasted divergence velocity in AeroFinsim with the properties of 6061 at current thickness. Perhaps I could make them a bit thicker? But then they add tailweight, which reduces stability.

Once I can get a STEP model for my IST, I can run that instant quote and have a final Yes or No answer on my IST. I agree, total cost needs to come down, so I will be seriously considering the costs when the quote comes back. If it is any more than what Im guessing, which by the sound of those with experience - it will be, then I will go back to the drawing board with the interstage and suspend the flight untill further notice. With each passing day, the chances of flying at Aeronaut grow dimmer and dimmer. To be honest, its a goal and thats about it anymore. I do not think it polite to waste TRA's time with a half baked design, so I am waiting untill I can really say Ive got a solid and fundable project before I submit for TRA Class3 board of review. This process is supposed to take 90 days. That puts me at XPRS already, at the earliest. Hopefully I will have something to move forward with by next week or perhaps next weekend.

I am open to everyone's ideas on how to design the interstage differently but here are some things to consider -
1. I would like to keep the full stack length as short and light weight as possible.
2. The advantage to having tongues that grip the outside of the fincan is that they allow the sustainer fins to be right at the ass of the sustainer, maximizing stability.
3. More moving parts allows for more failure.

Hmm.... Perhaps I could switch the Carbon fiber airframes to G12? Carbon is lighter, but G12 will knock airframe costs into a third. Little savings here and there could amount to a lot.. Just a thought.

 

[Martymonsta]

This is going to be an interesting project to watch.
I noticed that you are planning on using an CD3 as the primary separation method. Has something been done to rectify the mode of failure that was experienced by the AeroPac 100k project using CO2 for separation?

Also regarding the conversation about apogee detection. The main problem that I see with any accelerometer based system is cumulative error and rounding of data. A raven3 sampling at 200hz on a flight to 100k might have 20000 data samples and with a little bit of rounding, by apogee this had compounded into a massive error.

 

[butalane]

The professional motors you're referencing ran steel cases at pressures exponentially higher than what an aluminum cased CTI motor will run. Apples and oranges. Spend all the time you want getting your own waiver, knowing full well that three people (Jarvis, Anderson and Biba) who nailed similar flights got in the low 100s. If you execute everything flawlessly the first time (which none of them did, BTW) and get a similar result well below your 200k claim, prepare to be disappointed with a result you ought to be thrilled with due to talking the talk before walking the walk.

Have fun and see ya on the playa-

Steve

Actually the SuperLoki Dart case was Aluminum (2014 and later 2024 IIRC) and didn't run any crazy pressure -1500psi. Even at higher pressure the propellant in the first iterations of LOKI were low Al poly sulfide based that are not nearly as high performing as C* despite the vehicle optimization. Fun story - when the SL's were switched to HTPB propellant (with matched ballistic properties) they were still achieving higher performance than the poly sulfide equivalents.

The motors (especially CTI's) aren't the gating item to optimization in almost any of the high performance projects that are flown.

 

[UhClem]

Also regarding the conversation about apogee detection. The main problem that I see with any accelerometer based system is cumulative error and rounding of data. A raven3 sampling at 200hz on a flight to 100k might have 20000 data samples and with a little bit of rounding, by apogee this had compounded into a massive error.

Just where do you think this rounding error is coming from?

 

[BLKKROW]

I am happy to let everyone on TRF know that I will be doing the 3D printing and designing for the AV Bay sleds; I am offering my service to Ryan for free to help with the budget. I am really excited to watch this project unfold and fly! I know the 3D printing is not a huge expense but I figure I can help in a small way.

Ryan, have you talked to Eric from BadAzz in regards to the machine work? Maybe he can give you an idea of a way to change it to reduce the cost of the machining?

 

[A5tr0 An0n]

Indeed! One of my colleagues who was assisting me design the rocket was making models and running CFD in SolidWorks, but when he left the team I lost access to the software. I guess he lost his drive, and I agree it is an arduous process, but like it's been said.. Thats rocketry. The only CFD render I ever actually was able to secure is one showing dynamic pressure on the nosecone at high mach to justify placement of barometric relief ports and the viewing window for the GoPro. Cd values for the full stack could be obtained from RAS running a full stack profile, but lacking sustainer fins. I wish I had solidworks. All I have is AutoCAD. This being said, I am having my friend Chris C. help make updated models and perhaps he may be able to run some stress analy on the tongues and CFD on the stack, and I may finally get those renders. Depending on the resuslts, I may be able to run some lesser expensive materials for the fincan and IST, such as 6061 or something else. I have a strong feeling the fins will need to remain titanium, as they go far beyond forecasted divergence velocity in AeroFinsim with the properties of 6061 at current thickness. Perhaps I could make them a bit thicker? But then they add tailweight, which reduces stability.

Once I can get a STEP model for my IST, I can run that instant quote and have a final Yes or No answer on my IST. I agree, total cost needs to come down, so I will be seriously considering the costs when the quote comes back. If it is any more than what Im guessing, which by the sound of those with experience - it will be, then I will go back to the drawing board with the interstage and suspend the flight untill further notice. With each passing day, the chances of flying at Aeronaut grow dimmer and dimmer. To be honest, its a goal and thats about it anymore. I do not think it polite to waste TRA's time with a half baked design, so I am waiting untill I can really say Ive got a solid and fundable project before I submit for TRA Class3 board of review. This process is supposed to take 90 days. That puts me at XPRS already, at the earliest. Hopefully I will have something to move forward with by next week or perhaps next weekend.

I am open to everyone's ideas on how to design the interstage differently but here are some things to consider -
1. I would like to keep the full stack length as short and light weight as possible.
2. The advantage to having tongues that grip the outside of the fincan is that they allow the sustainer fins to be right at the ass of the sustainer, maximizing stability.
3. More moving parts allows for more failure.

Hmm.... Perhaps I could switch the Carbon fiber airframes to G12? Carbon is lighter, but G12 will knock airframe costs into a third. Little savings here and there could amount to a lot.. Just a thought.

View attachment 263952

Well that does suck and SW does not come cheap nor ANSYS. RAS for Cd is better than nothing, I suppose. If you were to pull this off in time for flight by the end of this year I think the best launch event would be BALLS seeing as they have had 300kft+ waivers and that would save a lot of time and headaches with trying to get your own wavier. That of course would require you to break your relationship with AeroPac launches but if everything else works out for you then this might be the best option to actually get it in the air.

I just read the ISC part of your write up and it appears your ISC design is to primarily minimize overall length of the vehicle. However I do think that to add ~3-6" to the overall length at the cost of saving thousands of dollars is a game changer. You can use the sustainer motor as the stack coupler and you will not gain much space over your current design as a result. Yes you will have some new design hurdles (stability one of them) but nothing to hard to design around. Any altitude losses due to this change would be "up in the air," with your predicted forecasts. Also if you want lightweight ( as you previously mentioned) go with composites for the material selection of the ISC, it can be done. On that note your sustainer fins do not need to be Ti and your booster fins are over designed being 0.157" thick plus the 10 layers (5 on each side) of tip-to-tip. My point? I think you have trade offs that you can do to save money, decrease time, and keep altitude to approximately the same but YOU will have to work harder with the changes.

I would keep CF as the AF material, I do not think the cost savings of G12 are enough for me personally to make that change. Have fun!