Workbench 2.0, Two-Stage 100K+ Build Thread and More

My first build thread on TRF! As winter fades I will be getting this rocket ready to fly once again and figured some of you might want to follow along and learn from my mistakes . AND KEEP ME FROM MAKING NEW ONES. I will add build pictures/write-ups as I have time.

How it started:
A few years ago I moved from Boston to Utah and decided to get back into flying given that Salt Lake is a half-day drive from Black Rock. Inspired by Aeropac's 100k flight in 2012 and that the largest motor casings I own are a 6G 98mm and 6G 75mm - I'd give my own interpretation of the 4" to 3" two-stager a try.

The rocket flew for the first time at Balls 24 but unfortunately the sustainer did not light for reasons I will explain later. Regardless, I recovered the stages in great shape and the camera payload worked as planned. Apogee was 31,000 ft and I'll chalk it up as a test flight.

Overall stats:
First stage
4" FWFG tubing
58" tall
7.94 lbs - dry (flight ready minus motor)
3 fins

Sustainer
3" FWFG tubing
5:1 VK FWFG
65.5" tall
6.98 lbs - dry (flight ready minus motor)
3 fins

Some of the goals for this project were:
- Vacuum bag all fin layups/reinforcements
- Avionics, payload, and interstage design/layout/machine code to be done in CAD (Creo 2.0) - in the case of a bad crash rebuilding/machining would not be a huge chore. Also helped me optimize layouts and test fit parts.
- Machine all G-10 bulkheads and av-boards using CNC
- 3D print some avionics components out of ABS
- Machine the aluminum transition of the interstage on the lathe
- Have two GoPros on-board. One taking video and the other taking timelapse pictures
- Control the 1-button on function of the GoPro using Arduino
- Design and build everything in the span of 3 months <- Go fever always bites!

THE FLIGHT!
First what I did wrong:
1.) Did not wire and e-match or setup a pyro channel in the event the second stage did not light. I added a 12 second delay on the sustainer primary apogee charges to account for exceeding the baro-sensor and accelerometer error during a nominal flight. A stupid mistake that cost me a 6" long zipper in the sustainer airframe.
2.) I selected the "flight state after" and "flight state before" radial buttons when programing the Easy Megas. DON'T DO THIS!!! This totally locked out second stage ignition despite meeting all other set criteria (velocity, angle from vert, altitude).
3.) I really rushed to put this project together by starting design/construction less than 3 months out from Balls. The month prior to launch I was already working every free moment I had and staying up until 2,3,4,5AM every night. If I had done a better job spreading this project out, I could've avoided mistakes #1, #2, abandoning my social life, loss of sleep, wasted N2501 motor etc. Lesson learned.

What went right:
1.) Perfect boost, rocket was 8 degrees off vertical when the sustainer was set to light at T+14 seconds and I could've let the sustainer coast a bit longer too.
2.) Recovery went on without a hitch (no damage minus sustainer zipper) with all charges firing and all parachutes deploying
3.) GPS lock maintained throughout the flight
4.) Payload performed really well - the Arduino activated GoPro worked perfectly. I got some great pictures and video from the flight.
5.) Rocket appears to have preformed better than simulated (RASAero predicted 28,000 ft for the sustainer no-light flight).
6.) The stack hit 1300 mph and the sustainer, acting as a boosted dart, flew to 31,000 ft.
7.) Low spin rate - good on board video
8.) 1.9 caliber stability margin, sustainer did not need any nose weight.

The photo uploader is not working great...so I will post what I can from the launch.





More to come soon!

Some more...and I'll post the flight video tomorrow.

Wow, nice project. Will be watching with interest!

Subscribed. Nice work!

I'd like to try a project like this someday when the kids are older and I have more free time on my hands. Until then I'll have to do it vicariously through threads like yours.

Flight Video!

Stats from Easy Mega:
Apogee: 30,786 ft
Speed: 1241 mph, Mach 1.6
Max G: 13

Motors: CTI N2501 staging to AT M685

[video=vimeo;160040633]https://vimeo.com/160040633[/video]

Next post: Fins

Looking forward to the build!

WOW this is awesome! Very pretty rocket! Even the less than perfect flight was still badass 30K aint bad.

So a 12 second apogee delay from apogee 'detection' , right? Was this to ensure that they fired on the way down no matter what? I had planned to do the same thing on a two stage of mine but now youve got me thinking--
if we're planning on blowing chutes at 100+ with no atmosphere... does it matter if it's still on the way up? I'd doubt 12 seconds would make that critical of a difference. You wouldnt be below 90K.
But clearly it DID matter that it was 12 seconds after it started the way down, given the sustainer was only at 30K.
Wow thats some great food for thought, I love it!
How do we combat this while maintaining the delay? I dont think the Megas have a failed stage detection mode.... boy would that rock!
perhaps the safest thing to do would be to just ditch the delay idea i guess... Man thats a good question.

Isnt it amazing tho, you still went 30K on a single mid N motor pushing hella extra weight? Badass stuff mate, Ill be watching this thread as it goes for sure!

I was really happy with the 30k flight. Makes me think that the RASAero2 136k sim is not in left field.

Ya I have no idea on the delay. I just guessed at 12 seconds. I know the guys who flew A Money Pit to 120k used a 10 second delay on their Easy Mega. Looks about right based off their on-board video.

Do I fly with the delay this year? I don't know yet so I'm in the same boat. This was my first flight with the Easy Megas so I will spend more time learning what others have done. I do know the Aeropac team was way late on their deployment for their one successful 104k flight and the rocket was just falling quite fast (1000ft/s??) with no issues.

Kip_Daugirdas said:

I was really happy with the 30k flight. Makes me think that the RASAero2 136k sim is not in left field.

Ya I have no idea on the delay. I just guessed at 12 seconds. I know the guys who flew A Money Pit to 120k used a 10 second delay on their Easy Mega. Looks about right based off their on-board video.

Do I fly with the delay this year? I don't know yet so I'm in the same boat. This was my first flight with the Easy Megas so I will spend more time learning what others have done. I do know the Aeropac team was way late on their deployment for their one successful 104k flight and the rocket was just falling quite fast (1000ft/s??) with no issues.

Click to expand...

You have a very positive outlook on things. I've never found much happiness when the sustainer doesn't light, although I can think of a few times when I would have liked to un-light a sustainer.

I'm confused about the 12 second delay. I know this type of approach has been suggested to compensate for early barometric apogee detection with the Raven (where the amount of delay to use depends on both the altitude and the speed). Is this why you used the delay, or did you just want to be late and not early? I can understand the confusion with the flight state (being as that still confuses me now).

Jim

op:

Subscribed!

JimJarvis50 said:

You have a very positive outlook on things. I've never found much happiness when the sustainer doesn't light, although I can think of a few times when I would have liked to un-light a sustainer.

I'm confused about the 12 second delay. I know this type of approach has been suggested to compensate for early barometric apogee detection with the Raven (where the amount of delay to use depends on both the altitude and the speed). Is this why you used the delay, or did you just want to be late and not early? I can understand the confusion with the flight state (being as that still confuses me now).

Jim

Click to expand...

The no light and zipper were a major bummer. Definitely left Black Rock with my tail between my legs. :-/

I saw your flight settings for the Easy in your thread (unfortunately after my flight). How did it go? Anything you would change for the next flight?

To answer your question, I figured some of the same issues that plagued the Raven above 100k were present in the Altus Metrum. And I definitely wanted to deploy late versus early.

Kip_Daugirdas said:

I saw your flight settings for the Easy in your thread (unfortunately after my flight). How did it go? Anything you would change for the next flight?

Click to expand...

The flight was quite interesting. All of the motors lit, but the results were otherwise disappointing. I had an EasyMega programming error too, but I lived through it. There were many lessons learned, and I hope to have more success going forward (two stage this year, maybe try the three stage the year after). I did a video of the on-board cameras below.

Jim

https://youtu.be/eHloNCGlYz4

Fins:

Not much new here. I borrowed techniques that others have used but I will add my two cents to the nuances of making composite fins.

Utah is great for a lot of things powder snow, skiing, mountain biking, rock climbing etc. But in terms of rocketry, the climate is good for doing epoxy work outside (May-September) in the garage - temps are tolerable and humidity is nonexistent.
Also my house isn't huge and I'd rather keep that type of work outside.

For the fin shape I find that it's a constant struggle between maximizing altitude and pushing the CP as far aft as possible. So I start out following some simple rules for minimum diameter rockets.
1) clipped delta shape
2) fin span = 1 airframe diameter
3) slightly forward canted trailing edge (so recovery cords can easily slide past fins)
4) at least 2X the span length for the root cord

Once the design is set it's off to cutting the cores. My cores for both the booster and sustainer are .050" thick G10. We have a proto shop at work for us engineers - no techs so it's up to us to learn how to use the equipment. It makes work a whole bunch more fun and we can use the shop after hours for personal stuff.

Oooo CNC fins. FYI never use an standard end mill for cutting G10 it will last for one set of fins. Use a cutter like this one: https://www.precisebits.com/products/carbidebits/fcrouter.asp



After the fins were cut they were surfaced prepped and laminated with 2 layers (on each side) of 3K (6oz) plain weave carbon at a 45/0/G-10/0/45 pattern. They were laminated on both sides (peel ply, perforated film, and breather on both sides too) and vacuum bagged to an aluminum plate that I milled flat (later would be used for bulkplates).

TIP 1: When laminating fins of any thickness (but particularly thin ones) do both sides at the same time! Otherwise you will likely be dealing with some kind of warp in the final product. Composites love symmetry at all times. Trust me on this one, I have plenty of warped laminated G10 plates which I'll probably turn into coffee table coasters.



The fins all came out with a .65 resin to cloth ratio and were .087" thick. I used HTR -212 resin. Which I have to say is awesome stuff. I like it better than Aeropoxy plus it has a high Tg.

Next came airfoils. I choose a 5 deg wedge - super sharp! I used the ghetto jig below and will definitely build a jig for a disk sander next time. Sanding 6 fins was two nights of work.


Then fins were then tacked on with JB Weld quick. Everything was kept in alignment using a jig I machined out of .25" MDF and two angle irons.


After that each fin got fillets of HTR-212 with milled fiberglass added.
TIP 2: I found that the ideal ratio of glass to resin is 1:2. I ran a dowel across each fillet maintaining a specific angle and wiping periodically.

However keeping the fillet shape perfectly consistent can be challenging with milled glass (I'm a perfectionist so it's prolly fine for most). After the fillets cured, I sanded them and did a once over with HTR-212 and fairing filler mixed in to make sure each fillet was spot on shape wise.


Part 2 will be vacuum bagging/finishing.

JimJarvis50 said:

I'm confused about the 12 second delay. I know this type of approach has been suggested to compensate for early barometric apogee detection with the Raven...

Click to expand...

FWIW, on our BALLS 100K+ attempt, I used two channels on the Raven, one for barometric apogee deploy with a max AGL limit in case the sustainer didn't light and one for inertial apogee plus a delta time in case it did (delta to make up for any accelerometer errors, rather be late than early), and wired them to the same ematch. Late apogee above 100K shouldn't matter because there's no air up there anyway. Of course none of this got tested because our booster motor blew up and trashed all the electronics anyway.

I hear you on the Easymega programming. Documentation needs some work IMHO.

Nice work, best of luck!

JimJarvis50 said:

I had an EasyMega programming error too, but I lived through it.

Click to expand...

Jim what was the programming error? Programming wise (assuming you are flying the Easy again) what would you do differently? Watching your video apogee detection was well timed.

BTW that was an awesome flight, it's a shame the stabilization module got jammed. You have had some great flights and more success than anyone I know flying above 100k.

Kip_Daugirdas said:

Jim what was the programming error? Programming wise (assuming you are flying the Easy again) what would you do differently? Watching your video apogee detection was well timed.

BTW that was an awesome flight, it's a shame the stabilization module got jammed. You have had some great flights and more success than anyone I know flying above 100k.

Click to expand...

As I recall, I intended to put in a delay after 2nd motor burnout for lighting the third stage. I also added an altitude check to inhibit the ignition if below a specified altitude. As you might expect, the altitude for the altitude check was reached after the motor burnout, and then that became the last of the "all other condition" that started the delay time. I think it ended up causing the motor to light something like 4 seconds later than I intended. The motor then took a while to come up to pressure, so the third stage lit quite a bit later than I planned.

Yes, the stabilization system not getting a chance was a bummer for the flight, but we did learn a lot about how it worked and have revised the programming and also a few of the mechanical aspects of the system. It should work better on future flights.

Jim

Awesome project man! I too have a second revision of a 4" to 3" 100K project that I hope to fly this year. My first attempt boosted out of the (short) tower at a severe angle; upper stage still lit without any off vertical restrictions from my Raven altimeters. I actually recovered the upper stage after it landed under chute a short 23 miles into the desert. I plan on using a long launch rail and a Telemega this go around...but now you guys have me nervous on the settings, when the time comes I might need a review of my second stage program logic from the veterans.

Burner said:

Awesome project man! I too have a second revision of a 4" to 3" 100K project that I hope to fly this year. My first attempt boosted out of the (short) tower at a severe angle; upper stage still lit without any off vertical restrictions from my Raven altimeters. I actually recovered the upper stage after it landed under chute a short 23 miles into the desert. I plan on using a long launch rail and a Telemega this go around...but now you guys have me nervous on the settings, when the time comes I might need a review of my second stage program logic from the veterans.

Click to expand...

Wow that's out there! I'm glad you found it, what did you use to track it? At what launch do you plan on re-flying?

In regards to the Altus Metrum gear it's really not that bad. Just every time you use new hardware you have to be super careful and I blew it in that regard.

I'm flying off a rail again for this attempt. The booster is used for 7 seconds and then cut loose. Drag from the rail guides doesn't have much of an effect on the overall flight.

I'm building a pad for this year's flight too. So if you are in need of a good pad with a 1515 rail, hit me up!

Kip_Daugirdas said:

Programming wise (assuming you are flying the Easy again) what would you do differently? Watching your video apogee detection was well timed.

Click to expand...

Forgot to respond to this part of your question ....

In my higher altitude flights, I just use timers for apogee deployment. This is combined with normal barometric apogee if the rocket reaches apogee at a lower altitude. I use a criteria of 90K feet for this. That is, if the rocket reaches a barometric apogee, and the altitude is less than 90K, then it's barometric deployment. Above 90K, this won't trigger deployment and deployment is instead based on the timers. This approach would have prevented your zipper.

I have discussed with Adrian the use of barometric deployment at above 100K. The problem is that the changes in the barometric sensor reading with altitude can't be distinguished from noise at some point. This point depends on both altitude and speed. At a given altitude, an altimeter might report apogee if it is moving slowly but not if it is moving faster. Thus, the approach is to let the altimeter determine barometric apogee and then apply a delay time to actually let the rocket coast to a stop. This approach doesn't seem much better to me than just using timers.

My preference would be to use inertial apogee. I know for sure, however, that the accuracy of this using Ravens is not sufficient. There are altimeters with better sensors where this might be possible, but I have not figured out a way to calibrate the accelerometer across its full range and then verify the accuracy. For a long coast period, and if not vertical, I suspect there would be issues.

So, until this can be done reliably with gps, I'm probably just going to continue using timers. The key to this is a good simulation and then being able to follow the plan. At least with RasAero II, we have a tool to handle the simulation part.

Jim

Kip_Daugirdas said:

Wow that's out there! I'm glad you found it, what did you use to track it? At what launch do you plan on re-flying?

Click to expand...

I used a 100mw BRB GPS unit, signal was loud and clear all the way till touch down. I'd likely fly it at XPRS or Balls.

I'm flying off a rail again for this attempt. The booster is used for 7 seconds and then cut loose. Drag from the rail guides doesn't have much of an effect on the overall flight.

Click to expand...

Completely agree, I came to the same conclusion. I was just using what I had available at the time.

I'm building a pad for this year's flight too. So if you are in need of a good pad with a 1515 rail, hit me up!

Click to expand...

That would be great, I'm hoping to avoid building my own pad but it sure is nice to be self sufficient out there.

Fins part 2 - lay ups

Again not much new here in regards to technique. This was my first time vacuum bagging a tip-to-tip layup and it turned out quite good. I used Jim Jarvis' tutorial. It's a time consuming process but well worth the results.

The layup pattern is staggered at 1/3 of fin area followed by 2/3, and finally full tip to tip. This technique helps keep the fin x-section thin and dampens out any vibrations.

The cloth used was 3K 6oz plain weave carbon. I chose it because it has a denser weave than twill.



.030 Thick at the leading, trailing, and edges


This is the final result. You can also see how I attached the rail guide on the booster. I used a PML urethane guide. I can't say I really like those...for this year's flight I might machine my own out of Delrin and replace this one.

Next up avionics/payload bay...

That layup looks beautiful!

Kip_Daugirdas said:

Fins part 2 - lay ups

Again not much new here in regards to technique. This was my first time vacuum bagging a tip-to-tip layup and it turned out quite good. I used Jim Jarvis' tutorial. It's a time consuming process but well worth the results.

Click to expand...

Kip, thanks for posting this. It makes having written the process down worthwhile. I just finished an application myself over the weekend, so I know the time you invested in your project. It looks like you got a great result!

Jim

JimJarvis50 said:

Kip, thanks for posting this. It makes having written the process down worthwhile. I just finished an application myself over the weekend, so I know the time you invested in your project. It looks like you got a great result!

Jim

Click to expand...

Jim I have to say thank you for taking the time to write the work instructions out and to such great detail! It made the process far less daunting! I can't believe how well the shelf technique worked. Thanks again!

I did not finish the leading edges with cotronics. I'm hoping that with the thin cross section of the fin and max velocity of the sustainer being around 30,000 ft aero heating won't cause any damage. The coating on the fabric was six thin applications of HTR-212.

Heat distortion temp of HTR is 290 F for those interested.

Booster avionics:

With this rocket I am working hard to keep the entire stack as short as possible. Components like avionics, recovery, motors are dictating the length.

Originally I was planning on using older avionics in the booster like my Adept ALTS altimeter, an ARTS flight computer, and some mini-Timer3s for stage separation. At best I could get this bay to be 4.5" long and sit directly above the 98mm forward closure. But taking a closer look at the forward closure I noticed there was a good deal of space between the tracking charge cylinder and ID of the airframe.

In that annulus, I could fit two Raven 3s and a Rocket Hunter transmitter with room to spare for batteries and wiring terminals. The forward bulkhead of the bay would attach directly to the motor casing. This solution was a huge space saver!

Here is the CAD:

Cross section showing the CTI forward closure


The mounts for the electronics were 3D printed and glued onto a 54mm fiberglass coupler which cleared the forward closure. The aft bulkhead was a .125" G-10. And the forward bulkhead is 6061 aluminum.

The forward bulkhead serves as motor retention and is held in place with four 8-32 machine screws that thread through the airframe. The thrust ring on the casing transmits the thrust forces/recovery forces into the airframe and not the 8-32 screws. Stainless steel 8-32 screws can handle about 262lbs in shear. So that's at the limit of four screws with a 800lb thrust motor plus 13Gs/rocket weight pushing the opposite direction.

The bay is then screwed to the forward closure with a 1/4-28 eyebolt that threads into forward closure. (The forward closure has a 3/8-16 to 1/4-28 threaded adapter in it so I didn't have to go overkill with the eyebolt).

Here are pics of the actual bay...

Heres a pic inside the tube that slides over the closure. The thing hanging inside there is a CR2032 batter holder which is used to power the Rocket Hunter transmitter. The transmitter has been rewired to use an on off switch.

Wow! Nice job on the av bay Kip.

ditto that... methodically planned and executed!
I'm taking notes for my 2-stage foray!

Very slick! Well done!

Looks great! Do you think heat from the engine would be a problem for the electronics (or possibly melting the mounts, if you made them out of PLA)?