(Yet Another) 100k Attempt

[robopup]

It's been a long term goal of mine to get something over 100k ft. I'd never felt quite confident that I had a reasonable shot a success until now. And, along with other legit prerequisites, watching a few Balls/XPRS attempts I'd always sort of assumed that the motor sizes required were something I wasn't that interested in dealing with, at least for now. But, inspired by a couple friends working on similar projects and, in particular, Kip's recent successful flight, I spun some sims this fall and realized it might be doable with a 75mm -> 54mm. I have enough worked out to post the summary below, but I'm sure I'm going to have a lot of questions: this is a lot of pretty new territory for me...

Edit: That got really long, my apologies. If you actually brave reading the whole thing, maybe some music helps.


Goals

• 100k apogee, recovery of both booster and sustainer
• Clear video of launch, through apogee, and ideally through sustainer touchdown
• As much home-rolled as possible
• (Relatively) low budget

Design Summary

• Two full stack rockets going into the season
• 75 mm booster -
  • 26.5 lb full stack pad weight estimate
  • Motor: ~M-1230, orange sunset, 8000 Ns, 6.5s burn time, 9.35 lb propellant, 15.75 lb loaded motor
  • 4 fins, > 3 calibers of stability through flight profile per Rasaero, 0.125 in G10 with no tip-to-tip
  • Structure - composite airframe, ~0.050 in carbon fiber (5 wraps) (plain weave carbon fiber with Adtech 820 from Soller Composites)
  • Electronics
    • single altimeter in ISC to manage dual deploy
• 54 mm sustainer:
  • 7.75 lb pad weight estimate
  • Motor: ~L-325, orange sunset, 2900 Ns, 9s burn time, 3.3 lb propellant, 5.85 lb loaded motor
  • 4 fins, > 2 calibers of stability through flight profile per Rasaero, 0.1 in G10 with no tip-to-tip
  • Structure - composite airframe, ~0.030 in carbon fiber (3 wraps) (same plain weave carbon fiber with Adtech 820 from Soller Composites)
  • Electronics
    • redundant in nosecone for recovery, one of which has telemetry and GPS
    • one to manage staging bolted to top of motor
  • 54 mm 5:1 von Karmen nosecone
  • Jim Jarvis style ignition via copper tape, for the full up arming test before sustainer ignitor goes in the motor. No head end.

Electronics

• Homerolled, I've exclusively been flying these the last few years in similar sized single stage projects
• Standard version has low and high g accelerometers, pressure and temperature sensors. Fancy version additionally has a GPS and radio for telemetry (I've been burned now 1 too many times with my Walston and other directional tracker; having time of flight only to never find the rocket after it's landed)
• User defined events
  • Up to 16 user defined events which can trigger pyro output using combinations of AND or OR logic
  • Triggered by liftoff, burnout, apogee, or any other user defined event
  • Greater than/less than/equal to constraints can be placed on altitude, velocity, acceleration, tilt angle, max tilt angle, flight phase, or powered flight counter. Additionally, time after triggering event and time after all constraints met can be specified.
• Simulation - not there quite yet, but I'm modifying post-scripts for testing I wrote a while ago to ingest Rasaero data -> convert to raw sensor output -> run on flight firmware in an attempt to remove surprises encountered by other folks attempting projects like this and to try to dial in my staging logic.
• Staging logic - TBD but I'm leaning away from a simple timer, and more towards a combination of:
  • min_speed < velocity < max_speed
  • tilt_angle < max_tilt and max_tilt_angle < max_tilt2
  • altitude > min_altitude

CAD

• Sustainer ebay:
• 
• Payload - staging altimeter and GoPro Session 5:
• 
• ISC:
• 
• Booster ebay:
• 

Flight Simulation
All simulations were done in Rasaero using weights determined from CAD. Chuck and Kip's new Rasaero thread on skin friction coefficients and simulation has been awesome... Goal is to keep the booster stability at > 3.0 calibers, sustainer at > 2.0 calibers through flight:






Test Plans
I have a number of test flights slotted for the spring; I want to make sure I feel super comfortable with all components before a full up attempt:

• 54 mm sustainer motor static test. My motor making buddies and I have quite a bit of flight experience with the booster motor, so not static testing that one.
• Sustainer single stage on a small H to test all electronics and cameras
• Staging test with small motors - ~I-600 in the booster -> small H in the sustainer
• Sustainer single stage on the full L-325 motor

Current Progress
Winters are usually pretty busy for me with my extra hobby job, but I've made decent progress on several components of this project which I was either more nervous about finishing in time or could be done before finalizing design

• Most of my time so far has gone into a ground up rewrite up my electronics firmware which I've just about finished. Everything is much more modular, I can define the user events described above, I added support for telemetry with a separate ground station I've been working on along with GPS and I have a full suite of unit and flight tests I've been meaning to write for a while. Big remaining item is to finish the scripts to ingest simulation data for full simulated flights
• 8 of 12 75 mm booster bates grains are cast (going for 2 booster motors total). I have my sustainer motor designed, and just took delivery of some of the fancy pants Loki liners but have not cast grains yet. I need to cast 4 sustainer motors: one for static testing, one for a single stage flight, and two flight ready motors.
• Tubing - I have all tubing rolled for two each of: 54 mm sustainers, ISC 54mm and 75mm and 75 mm boosters
• 
• Fin beveling jig - I built a v2, trying to fix everything that was bothering me from v1 https://www.rocketryforum.com/threads/75mm-composite-minimum-diameter-build.173626/post-2312429. I'm sure I'll find new stuff I'm not happy about with this one
• ISC - At some point I'll write a post detailing my intended design, but for now the 3d printed portion of this part has gone through a few design iterations with some cheapazoidal leftover material:
• 

Game on!

 

[mrwalsh85]

Game on, indeed!

Love your CAD work. I've been trying to find the motivation to challenge myself to get more compact with my electronics, just as you did. I do have some minimum diameter projects in the pipeline, so maybe that'll be a good test bed. Following!

 

[waltr]

Wow, great write up.
This is way beyond anything I could do but will be following this thread and your progress.

All the best luck.

 

[Kip_Daugirdas]

Design looks really good! Thanks for sharing the project, there haven’t been too many like this on TRF in a while.

I’m not sure what it would buy you but expanding out the nozzle on the second stage might gain a bit of altitude. Looks like with a phenolic carrier doubling as the expansion cone the case doesn’t have to get any longer. Phenolic will also help keep that case a bit cooler during the long moon burn. But what you have is great too!

Here’s a 54mm (left) with a 75mm (right)

 

[robopup]

@Kip_Daugirdas Those do look sweet... I've never used phenolic in my nozzles before. Do yours work like the typical aluminum designs, where the graphite is inserted with it's own orings, or is it all somehow bonded and machined as one piece?

The 54mm nozzle expansion is currently optimized for ~22k ft. Designing this motor, I've been appreciating orange sunset's high Kn's even more than usual - the extra expansion still fits in my standard full diameter 54 mm nozzle size. Probably I'll open it up a little more, to the diameter of the thrust ring, because I want enough width on the thrust ring for a aft plug I've been noodling on to help with sustainer ignition.

 

[Kip_Daugirdas]

Do yours work like the typical aluminum designs, where the graphite is inserted with it's own orings, or is it all somehow bonded and machined as one piece?

It’s a two step process I machine the phenolic carrier OD, o-ring groove(s), and nozzle step. Then I cut the taper for the graphite throat. Right after I remove the phenolic from the lathe and machine the graphite in the same setup. Then I JB Weld the graphite throat into the phenolic and leave it clamped overnight. Come back the following day after the glue has dried and finish up the nozzle entrance cone, throat, and exit.

Here’s a cross section of the carrier this is for a 4.5” motor but the design is identical 3” and 2”. Just make sure to buy phenolic that is roll wrapped not stacked layers. Typically you can find what you need on eBay, otherwise I special order a stick.

Hope this helps!

 

[rharshberger]

Good luck Robo, let me know if there is anything I can do to help!

 

[mrwalsh85]

Pretty much what we did for Marvin's O motor as well.

We used stacked phenolic I believe. Worked fine. We also RTV'd the insert in place. I believe the insert came out when the motor was cleaned up. It's been reassembled already and loaded for next flight.
I enjoyed machining this one. Took my time as I didn't know how the phenolic and graphite would react to machining but did just fine. Had someone else help hold the vacuum while I did the machining.

 

[robopup]

@Kip_Daugirdas Thanks for the explanation and drawing, that makes sense. Am I on the right track with something like this? Among other sustainer motor improvements, I would eventually like to design a setup to hang at least some of the expansion cone out the back, similar to your setup, and then have a tailcone which integrates over the back. I'm trying pretty hard to stay focused right now on getting a simple and familiar (to me) initial design up and working, before I start making incremental improvements to it. I'll fire this sustainer static test as designed and see how things go first.

@rharshberger - thanks Rich! I'm hoping to make pretty heavy use of your guys waiver this spring. I'll try to keep the fiddle factor to a minimum

@mrwalsh85 - dang, that nozzle is sexy too, thanks for the picture.

 

[Kip_Daugirdas]

Am I on the right track with something like this?

Yep, LE phenolic. Have the carrier designed before ordering so you get the right wall thickness, I made that mistake. I think I ordered 2” OD 1” ID for the 54mm nozzles.

 

[DeeRoc29]

Just make sure to buy phenolic that is roll wrapped not stacked layers.

Really? I'd think the roll wrapped tube type would erode faster than a carrier machined from a stacked block of phenolic. I've got a couple 6" cubes from Boedecker waiting to become carriers for Q motors.

 

[eggplant]

Really? I'd think the roll wrapped tube type would erode faster than a carrier machined from a stacked block of phenolic. I've got a couple 6" cubes from Boedecker waiting to become carriers for Q motors.

It isn't about erosion, it is about being structurally up to the task. I'm sure you can make a motor work with it, but I had a very negative experience with 4" phenolic LE round from McMaster (after 3" had worked just fine). This stuff is turned down from a super thick sheet with the layers oriented parallel to the diameter of the round. Best I can tell, the pressure in the divergent section cracked it almost immediately on the motor coming up and then exhaust went through the crack, widening it. It could have been due to lower phenolic content because the round was a much lighter color than the 3" I had used, but I'm not sure. Meanwhile, I've had the standard stacked-disk design withstand well over 1000 psi in 6" motors.

 

[robopup]

Probably one of those things where either work... up to a point. I figure if I was choosing I'd prefer it rolled over stacked for the stiffness and strength of the fiber over the matrix. Having said that, I'm not very familiar with the constituent fiber properties. Either way, something for the future, at least for me...

 

[Sandy H.]

@robopup I am zero qualified to help in any way, but I will read and enjoy, so please post any and everything you think would interest the members.

I look forward to seeing your successful flight(s). This kind of rocketry is astounding to me and it is awesome. I hope to see some flight this extreme in person during my lifetime.

Sandy.

 

[Kip_Daugirdas]

Really? I'd think the roll wrapped tube type would erode faster than a carrier machined from a stacked block of phenolic.

I don’t want to turn @robopup thread into a nozzle discussion. But here are a couple photos. I’ve never done the stacked block method so I cannot provide a comparison. Perhaps these posts can get moved into the research forum?

Ground test 4.5” nozzle (cracking happened after firing during cool down)


4.5” flight nozzle

 

[Grog6]

MDF is masically like a PC board, IIRC. It's glass fibers embedded in a matrix of epoxy. The epoxt can be varied to supply any needed properties, like phenolic resin, for example
To know how strong it it, you need to know what it's made of.

 

[robopup]

Design is finalized, at least enough for me to actually get around to doing some version control and feel good about cutting fins. I did a full audit of all dimensions, densities and masses in CAD, and a review of both motor designs in burnsim. Hopefully I didn't miss anything too important... There were some minor corrections from my above estimates:

• largest mass change - I had not included the weight of the recovery gear in the booster
• Slightly shorter sustainer casing to account for as measured length of my liners (they're expensive so I want to get two motors out of each)
• increased the nozzle expansion ratio in the sustainer such that it is the diameter of my stock thrust rings
• increased the nozzle expansion ratio in the booster to size for ambient pressure at Black Rock. Probably not optimal, but close enough for my purposes

After making the above mass changes, I went through several iterations of tweaking fin spans in CAD -> pulling mass and Cg to Rasaero -> simulation, until full stack stability was > 3.0 calibers and sustainer stability was > 2.0 calibers (sustainer is actually at ~2.2, I gave my self a little wiggle room in my constraint because of the variation due to speed). This resulted in slightly larger fins (0.25 in for the booster, .125 in for the sustainer) to account for the above bulleted changes.

Here are some final numbers on my masses and motors:




Next, per the discussion in the Mesos/Rasaero thread, I attempted to bracket expected altitude for a nominal flight with variation in skin friction drag. I set skin friction at each of rough camo paint, camo paint and smooth paint then iteratively changed sustainer ignition delay for each until sustainer ignition occured at mach 0.9. I'm not totally set on this number (my original sim used something like mach 0.75), and I'd prefer to coast a little longer if I'm not tilting, but it seems like a reasonable starting point. Here's what I got out of that exercise:


I think I'm going to sleep on it one more night, and then cut fins. I feel about like this sunrise from last weekend - I'm pretty sick of my computer at this point and excited to get on with the actual build!

 

[robopup]

I had a typo in my above motor spreadsheet - here's the correct final design data:

Moving on to fins. I've got all 4 sets of 4 cut on an old tile saw and sanded even. I'm using 0.1" G10 for the sustainers and 0.125" G10 for the boosters. I've also started working on the bevels. I've been messing around with jigs for a few years now, and I think I've finally landed on something I'm happy with.

I used a piece of plywood as a base with a few 3d printed pieces which match the moat in the saw to hold things in place:


I have a couple more pieces of plywood held at an angle by a couple more 3d prints - I'm using 7.5deg. The back of the jig presses up against a fence, and some extra G10 bolted to the front holds the fin in place. The plywood is sealed with a couple coats of laminating resin and waxed so that things move smoothly - I tried to cut this particular corner at first by not sealing and botched my first fin because things didn't slide smoothly enough.






First set of sustainer fins is finished. It's still too cold here to be messing with a wet tile saw for long before my fingers become useless, so I'll do the remaining sets the next few days.

 

[robopup]

Tacking on fins

For any structural bond I care about, I basically follow the steps in the excellent stickied thread here. This is what my process looks like on my sustainer fincan:

Approximate fin locations marked with a sharpie so I know where to sand:


Sanded tubing until I had a water break surface. Fin roots are hard to tell, so I usually sand them for a while and call it good:


Wiped clean with acetone:


Guides in place. These are 3d printed pieces with the same thickness as the fin material which keep things orthogonal and straight (hopefully). Rings at the top and bottom space fins at 90deg.


Pieces of angle hold things in place while the epoxy cures. It looks pretty ridiculous but I've been happy with the results so far.Lately I've been using E-60HP (or 120) for all structural bonds and fillets that I care about.


Both sustainer fincans have fins on. Here's one completed. Onto booster fincans and fillets now...

 

[robopup]

A few updates:

Altimeter firmware testing:

I've been a little spooked by all the issues documented on this forum getting altimeters programed correctly for higher altitude flights like this, so I have a requirement that I need to be able to run simulation/flight data through my altimeter firmware to test that the logic I use for pyro events is correct (and obviously for bug testing also). The goal is to be able to compile my firmware code for testing and debugging on my laptop which is as close as possible to the exact code I'm running on my altimeters. The way I have things set up is:

• I build and upload firmware to altimeters via PlatformIO (I run Atmel samd chips with an Arduino Zero bootloader) OR compile with a python wrapper to run test scripts
• All firmware code is the same except for sensor drivers:
  • firmware sensor drivers (along with other hardware) for accelerometers, barometer and GPS are in their own folder - these grab raw sensor output (usually 2 byte or 4 byte numbers for my stuff) and do nothing else.
  • replica sensor drivers for testing ingest sim/flight data and convert this data to raw sensor output
• In the test version, I have a for loop which reads sim/flight data, passes it to the replica sensor drivers then calls the firmware loop
• my compilation process uses the appropriate include folder from above to gather drivers, and compiles the rest of the firmware untouched

The plots below don't look like much but I'm pretty excited - the first passes altimeter data from a 54mm flight I did last fall through the process above, the second passes Rasaero simulation data of the same flight through. There are a couple other minor items I need to take care of before I can actually use this for my two stage project, but I'm just about there.

The altimeter configuration I'm using is also shown below for reference (I have three output channels, but I can apply up to 9 user defined events on each channel with OR logic to fire the channels - for use cases like whether to fire drogue using a timer because of successful sustainer ignition or to use standard apogee detection if sustainer doesn't light). These are simple right now, just my typical dual deploy setup:
*Event,Active=1,Channel=APOGEE,Event=APOGEE,DelayAfterEvent=0,DelayAfterConstraintsMet=0,Constraints=None
*Event,Active=1,Channel=MAIN,Event=APOGEE,DelayAfterEvent=0,DelayAfterConstraintsMet=0,Constraints=ALTITUDE<500000
*Event,Active=1,Channel=THIRD,Event=APOGEE,DelayAfterEvent=1000,DelayAfterConstraintsMet=0,Constraints=None
*Output,Active=1,Channel=APOGEE,TimeOn=250
*Output,Active=1,Channel=MAIN,TimeOn=250
*Output,Active=1,Channel=THIRD,TimeOn=250

Flight data:


Rasaero data (I obviously didn't add a chute deployment to the sim here):


Altimeter hardware:

I've also made a lot of progress on updated hardware. Last fall I lost a rocket that only had RF tracking (but good time of flight); that frustration got channeled into a telemetry+GPS version of my altimeter. I've bench tested all hardware, have good GPS data and am able to transmit data packets to a separate ground station. Next steps are to take a few bike rides to play with the radio range and settings and to figure out a simple Android app to view packets.




Fincans:

I've got all the fins tacked on. I haven't done any fillets yet - the weather hasn't been great for prep sanding here and I'd prefer to keep the dust out of the house so I'm going to give it another week and see if I get a couple nice days.

 

[robopup]

Sustainer electronics mock-ups

It usually takes me two or three shots to dial in all the tolerances on printed ebays. I'm getting close on the sustainer bays.

This will go in the nosecone. It controls primary and backup deployment charges, and will provide GPS tracking:


Battery side - These are all single cell 250 mah. Telemetry altimeter will get two of these batteries, regular will get one.


I run my altimeters with kalman filtered baro and accelerometer data. I was a little nervous when I started mounting altimeters this far in the nose with vent holes, but have several successful flights under my belt now and feel pretty good about not getting any false positives due to vent location. Here's a plot of a 54mm flight I did last fall comparing accelerometer based altitude, baro based altitude (very bad data through mach - altitude even goes negative!) and kalman filtered altitude (calculated on my altimeter in situ):


The other sustainer ebay will get bolted onto the front of the motor. Its altimeter controls sustainer ignition and also holds a Gopro Hero Session 5. The three screw switches are for:

• altimeter power
• sustainer igniter open/close switch between altimeter and igniter
• shunt on igniter

Here's a quick circuit diagram of my plan for the switches and the sustainer igniter:







The last few days I've been taking a closer look at windows for my camera - this is the first time I've done on board video, there's been a lot of reading on this site and a bit of a learning curve going on. I've convinced myself I want a window in the first place, the hole is big enough that I'm nervous about messing with the baro sensor on the altimeter in the bay. I do have a couple questions for the group I haven't been able to find satisfactory answers for:

• How necessary is sapphire glass? It seems to be the preferred material for these types of flights, but I'm still wondering if I could get away with regular glass or a plastic like acrylic; So far I haven't found a demonstration that these cheaper materials don't work. Anyone aware of a successful video capture without the sapphire?
• If sapphire is necessary, any thoughts on the cheapest place to buy small circles? I've been astonished by the prices for 15mm diameter x 0.5mm thick windows at places like: https://www.edmundoptics.com/f/sapphire-windows/12234/ (I believe I found this link from @FredA somewhere).

 

[OverTheTop]

I ran into this website a while back (somebody else here found it) and bookmarked it. I suggest you do the same .

https://www.esslinger.com/watch-min...rved-rectangular-1-5mm-thickness-for-cutting/
Great for curved camera windows for airframes.

 

[jderimig]

You don't need sapphire. If you did, your fiberglass wouldn't be good enough either.

 

[robopup]

@OverTheTop - That link is excellent, thank you! Just curious if you've flown this stuff? EDIT: I believe this is the thread the link came from, with video of the launch, https://www.rocketryforum.com/threads/l3-project-modular-98mm.160518/page-2, along with how to calculate the correct radius here: https://tutors.com/lesson/how-to-find-arc-measure-formula

@jderimig - my thoughts too - really the only thing that gives me pause at this point is that other folks doing high/fast projects would be willing to pay $75 for a quarter sized piece of glass.

 

[FredA]

I forget where I found them, but ended up more like $6 each.
Need is more around turbulence....if' you're flying hard enough to melt fin leading edges you might need something tough if your window causes lots of turbulence.
AND, we were building a Mach-4+ camera platform, why skimp on the window.

 

[robopup]

Thanks Fred. I don't mind spending on hardware; I just want to make sure I have a clear and demonstrable reason why. At $6 (or ~$15 on OverTheTop's link) vs $75, I'll probably just order a sampling and not overthink it. Starting to look like I just didn't google quite hard enough for a reasonable site to purchase...

 

[prfesser]

Edmund's sapphire windows are intended for high-precision optical work, with special coatings, etc. And IMO Edmund today is rather overpriced for most amateur uses. If a material is needed that is transparent, very flat, and can withstand high temperatures and rapid temperature changes, fused silica (fused quartz) should be adequate. Rather less expensive than sapphire.

 

[robopup]

Lots of loose ends starting to get tied off here which is exciting...

First up - thanks again for the suggestions on glass; here's what an 18mm x 0.8mm glass cover from Esslinger looks like:


Plan is to epoxy this into the payload coupler tube like this:




All fillets on the fincans are also finished. I used EA E-60HP on all with the same process as described here. The boosters got 0.75 in radius, sustainers 0.5 in. My plan is still to omit tip to tip layups so other than finishing work the fincans are complete. I'm happy with final weights compared to CAD - after paint they should be right around the predicted values:

• Sustainers
  • CAD - 270 g
  • A - 221 g
  • B - 221 g
• Boosters
  • CAD - 865 g
  • A - 797 g
  • B - 799 g

 

[robopup]

Nosecones

Both nosecones are also finished. I spent years fiddling with overlapping seams on molds, but was never quite happy with them - they were definitely the weak spot in my cones and the internal thickness change made it hard to nail 3d printed ebay sleds. For the last year or so I've been playing around with internal liners with Soller Composites sleeves and have been really happy with the results.

Liners - cylindrical section is 1 wrap of 6oz plain weave carbon fiber. Curved section is 3d printed:


Superglued together and sanded:


Fabric gets laid up onto the liner:


Butt splice where carbon ends and fiberglass begins, so my gps tracker will work:


All fabric on:


Mold painted with a couple layers of PVA. This thread shows how I made a similar 75mm mold if you're interested. I place the nosecone in the mold, backed out from where it should be by about an inch. Then close the halves, tighten the bolts and push the cone in as far as it will go:




This is baking in my oven now - I do ~8 hours at 40 C, then post cure at 75 C for a couple hours. Here's the first cone, trimmed and with screw switch holes drilled. Final step will be to cut off the nose tip, to be replaced with aluminum.

 

[tfish]

Next level **** going on here!

Tony