L3 project build thread

[caveduck]

After 40 or so years in rocketry I think it's finally time to go start on my NAR L3 cert. ...(takes deep breath)... In this initial post I'll lay out my thoughts on the project tradeoffs, risk control, and constraints that I have to work with.

First the basic constraints -

• We have a great flying site at Plaster City with 25kft waivers and onsite vendor
• The desert surface at the launch site is pretty unforgiving.
• Available transport vehicle is a medium SUV
• Don't have a large shop with advanced fabrication equipment
• Need to be efficient with my available building time

The point of an L3 project is to do a carefully documented, high-reliability flight using a Really High Power(tm) motor that meets certain technical requirements, primarily having fully redundant electronic recovery deployment, and being recovered without damage that would affect flightworthiness. As a longtime engineer, I know that this translates directly into controlling risks as much as possible. The types of risk in an L3 project are

• Design risk
• Construction risk
• Flight risk
• Recovery risk

Design risk
Though I could certainly do my own design (having once written a 6-DOF flight sim), reliability will be maximized by using a well-proven existing design. This is a basic make/buy type of decision. Objectives here:

• Use an all-fiberglass vehicle; sturdy, no flight dynamics issues and will resist landing damage
• Keep flight speeds below Mach 0.8 or so; no point in getting into high-Q flight regimes.
• Keep the vehicle on the lighter side (10-12 kg launch mass) - reduces kinetic energy delivered at landing.
• Keep the motor on the small side - holds down flight speed and cost
• Use very well-known and proven electronics modules

Construction risk

• Avoid exotic or really time-consuming construction methods. In particular I do not want to do a lot of custom fiberglassing.
• Use high-grade adhesives appropriate to the purposes, e.g. West Marine, System 3, JB Weld
• Static test structures (esp. fin attachment) to reasonable loads.
• Would ideally like to get the project done by ~June to allow time for other things during the LDRS/NARAM/BALLS season

Flight risk

• Ground test deployment subsystems to determine correct BP charge mass and verify e-match batch reliability
• Debug the checklists thoroughly. Then debug them again.
• Flight test the vehicle, launcher and all electronics with K motors to smoke out any ops issues before the cert flight
• Have secure position retention for the electronics switches
• Use Cesaroni motor for simplest assembly (but see if there is contrary reliability data)
• Have a backup vehicle and spare (tested) electronics. This is also a schedule risk mitigation.

Recovery risk

• Use a GPS transponder (e.g. Beeline) to prevent loss
• Keep altitude under 10kft; 5-7Kft would be preferred (but may not be possible with smallish rocket)
• Size parachutes and time the deployments to keep flight duration to 3-4 minutes
• Don't skimp on the main chute size, you want low landing velocity
• Low aspect ratio fins greatly reduce damage potential

First Steps and Decisions
After looking at available motors, reading various L3 build blogs (some with really huge construction efforts), surveying some vendors and talking to our local L3CC member I decided to start with a largish-but-not-huge all-fiberglass kit with a 75mm motor mount supporting plenty of K/L/M motor choices. A lot of them have split fins now which I preferred to avoid (added build complexity and individually higher aspect ratios). KenRico's recent flights showed that a Vindicator will do ~5500 on a fairly full K, so something in that size range ought to work fine; going to a low L should stay well under 10Kft. The cert flight will be higher but there's not much I can do about that. After that I homed in pretty quickly on the Madcow DX3-XL, which has nice low aspect ratio fins, ~10kg liftoff mass, reasonable cost, will fit in my car, and should have moderate construction time, allowing me to build a 2nd flight article without requiring an extra 6 months.

(finally a photo!) I picked up the first DX3-XL at a recent launch. Here's the obligatory parts-inventory photo


The other day I grabbed the Madcow RockSim file for this design and quickly found that it's riddled with errors. I fired up OpenRocket and got most of the problems straightened out in the last couple of evenings - when it's ready I'll post it. The altitudes look like 5900' with a K555 and 7300' with an L820 (max Mach 0.75) so they're squarely in the desired flight regime for the test campaign. The sim file will be used to generate the required scale drawing and flight parameters predictions.

I'm deferring for now any final decisions about the parachutes and electronics. One of the altimeters will probably be a Raven and there will likely be a GPS transponder up in the nose...the other deployment computer is undetermined. I'm probably going to standardize on a 2.0 to 2.25" wide sled for cross vehicle compatibility and I'll need to make sure that everything fits comfortably (the Raven is a big help though). If it works out and looks reusable enough I might make the sled into a simple PCB to hold down the wire clutter and provide soldered-down connectors. I already made a general connection schematic for the electronics too.

More later, that was kind of a major braindump there.

 

[fyrwrxz]

Right on, Dave! Nicely thought out reasons and parameters for your choice. Looking forward to seeing it in person when that day arrives. Any help you may need is only a PM or call away.

 

[EeebeeE]

I can't tell by reading your post whether your DX3 XL is going to be used in your cert attempt or not, but according to the NAR Certification requirements for L3:
"The rocket shall use a motor with total impulse greater than 5120 Newton seconds"

This means your cert attempt must be with an M or greater. I suggest you use a CTI M1230, which is barely and M and is often used for L3 attempts. Although with that rocket design (assuming you are using the 4"), an M1230 is going to get you up to about Mach 1.3 and take you up over 17,000'. The altitude is within your waiver and I am pretty sure that particular rocket will survive mach speeds and it is better to punch through the transonic speed range than to play in it for a couple seconds. This rocket seems to be able to do that, but if you want to stay below mach, you may want to consider 7.5" in diameter or greater.

Re: Design risk. Another option is to upscale a proven design. I am looking at building a 3x upscale Estes Der V-3 as my L3 attempt and another person in our club is in the process of upscaling an Estes Cherokee. It involves fabricating your own fins, but that isn't too hard. I have stack-sawed plywood and fiberglass and get good uniformity in shape. It's even easier if you have or know someone who has a CNC cutter.

As far as electronics go, a Perfect Flight Stratologger is a very reliable and inexpensive altimeter. It automatically compensates for pressure changes as you go into mach speeds and will work up to 100,000'. Other than because of human error (personal experience ), I've never seen one fail. Great devices for approx $70 each. Well built and if you buy the USB attachment device, you can look at detailed telemetry data on your computer.

If you aren't going to go much past 10,000' the Beeline x-mitter is also a great choice. The benefit of that is that it transmits for days under a single charge, easily fits into a nose cone, and most likely several people with trackers are using the same thing so you can have multiple antennae monitoring your flight which gives you an opportunity to triangulate the descent and find your rocket faster.

BTW, if your are building another rocket for your L3 attempt, it would still be cool to fly a hefty M in that DX3 XL and let her rip. Now you see it...now you don't.

 

[caveduck]

Hi Evan, good observations all the way around...as you point out the actual cert flight is going to be higher and a bit above Mach 1, but it's not problematic for our waiver and the vehicle will take it without any real difficulty. Ideally I'd like the altiitude for the cert flight to be somewhat less, but at this point the space-related issues of building and transporting a 6"-7.5" rocket are pushing me pretty strongly to keep the size down and deal with the altitude. Actually anybody that knows me will realize that I REALLY like altitude, but I have to sit on my hands somewhat for this effort! :bangpan: There are reasons I want Ravens in my inventory (minimum diameter moonburners anyone??) The Stratologger is definitely on the shortlist for the 2nd altimeter.

Even with a 15-17Kft flight the landing should be < 1 mile from the launch site - likely could do OK with the 900MHz but I imagine I'll want to go get the Technician ham license so I can use the high power 70cm transmitter and be really sure. https://www.thefintels.com/aer/trackingbeacon.htm gives some test results showing detectability out to at least 3 miles even when non-LOS with that transmitter and a decent yagi.

BTW on a side note, the Beeline site bigredbee.com is currently flagged by McAfee for suspected malicious behavior on one of its pages and is also on some blacklists. I wonder if their web server's been compromised.

 

[troj]

BTW on a side note, the Beeline site bigredbee.com is currently flagged by McAfee for suspected malicious behavior on one of its pages and is also on some blacklists. I wonder if their web server's been compromised.

It just scanned clean.

https://sitecheck.sucuri.net/results/www.bigredbee.com

What're you seeing? Neither Firefox nor Chrome are complaining about it being on any blacklists.

-Kevin

 

[caveduck]

My (current) McAfee got a malicious behavior warning hit last night on one page (unfortunately I don't remember which), and the site is also blacklisted in my (large enough to be public) company's web gateway, cited for potential malicous content.

 

[bigredbee]

If you find a current problem that Mcafee points out, I'd appreciate the details if you have them. I know there are some https pages that contain insecure content (like the counters, etc) that can sometimes throw errors, but that's not malicious.

A long long time ago, I did run a mailserver, so on the quick scan of blacklisted sites I did, bigredbee.com showed up on 3 of the approx 40 sites as blacklisted due to email issues. I no longer run the mail server.

Please contact me off list greg at bigredbee dot com if you have any concerns

I've run all the virus checkers I've got (norton, spybot search and destroy as well as the engines on the web that assert PCI compliance for credit card transactions) and they all say I'm clean

Now -- back to the L3 build thread

 

[caveduck]

Thanks, glad to know that folks are listening! I've had similar sporadic issues on caveduck.com due to shared-hosting IP addrs getting blacklisted. Usually the hosting provider gets it cleared up quickly. Not sure why the current McAfee whined, will re-test tonight. If it's still not happy I'll email you the particulars (along with my order :wink

 

[troj]

BTW, if you've never build one before, MadCow does a really nice job with their kits.

Looking forward to seeing you progress on this!

-Kevin

 

[jcowles3]

I just was at plaster blaster and launched my L3 frenzy XL... same rocket or there abouts. here is the raven 2 dataif you want more info I'll be glad to help...

 

[caveduck]

I worked some more tonight on fixing up the Madcow-furnished DX3 XL sim file. Since some of the individual items didn't have believable masses I went over to Bed Bath and Beyond to pick up a kitchen scale. They had a good selection; I picked a Cuisinart model with 25lb capacity -- twice that of most of the others -- and 1 gm resolution. I did a quick checkout on the scale and found that it has a minimum threshold of around 10 gm (no surprise), but does have 1 gm sensitivity above that. With it I weighed all the important parts, measured key dimensions on some of them, and updated the simulation file.

The .ork file is really getting into decent shape now. The current empty mass is within about 30 gm of the override mass that Madcow had, though I haven't yet allowed for the epoxy in the motor mount / fin area. That seems OK since I've got some net additional mass over the baseline design by putting a telemetry sled for the Beeline in the nose cone coupler. I've attached the now heavily hacked .ork file and a screenshot of the drawing it generates just so you can see its current state.

With this file the postulated cert flight on a M1230 (4-grain) comes out to 14kft, max speed of 1.25 Mach and acceleration of around 14 g's. A tall flight, and a bit faster than I wanted, but not ridiculous and not pushing into high-risk terrain. I still need to finalize the masses of the avionics and do something to predict the net rearward CG shift coming from the epoxy. Final results will have to wait for the first vehicle to be built but in the meantime I can play with some arbitrary mass items. The ~300 gm that is probably still missing from the aft end reduces the alt 100' to 13.9kft, speed to 1.22Mach and leaves 1.47 cal stability.

Meanwhile my Madcow Black Friday shipment showed up with the pair of 3"/38mm Aerobees and the 29mm Bomarc. Oh Yeah!!

View attachment Madcow_DX3_XL_dbc_r1.ork

 

[caveduck]

After reading through the Beeline descriptions a bit more I ordered a 900MHz spread spectrum Beeline system for the first go, keeping the ham license requirement from getting in the way and not having to go separately get a Yaesu handheld etc.

 

[cwbullet]

Dave,

Are you an engineer? That is a slick write up in the first post. I have only seen that in the military and with engineers.

 

[caveduck]

*cough*...guilty as charged, yer honor.

 

[troj]

FWIW, while you can enter all the various bits and their proper weights, it's a lot easier to let it approximate the weight, then override it at the end, after it's built.

-Kevin

 

[Quaranta]

What happened? Looks like the main didn't deploy properly.

Agree. The file title is "JWC Cert Flight Level 3 FAILED".

Charge fired (obviously), but only very subtle change in descent rate. Tangled main? Compartment separation w/o deploy perhaps?

 

[jcowles3]

Winner... Failed is correct, main got melted off the spinner by the deployment charge, deployed but just kept floating away! Landed with little damage to the wood slead but nothing else was hurt, love the landing in the file. But the post was added for caveduck to see mach on an M2250-C star... Love my Raven2

 

[jcowles3]

One addition thought is I mounted my upper rail guide as madcow placed it (assuming yours is similarly placed) and feel it's too low and allowed my rocket to have some nice rail rash... I have thoughts of moving mine up to prevent it...

 

[caveduck]

FWIW, while you can enter all the various bits and their proper weights, it's a lot easier to let it approximate the weight, then override it at the end, after it's built.

True enough, but now I have a reasonable "live" simulation with which to see altitude and CG change effects long before actually building anything. Also I wanted to verify their CG and didn't really trust the file because the subcomponents in it didn't match the as-delivered kit. With a global override in place you are pretty much in the dark on CG shifts. Anyway it didn't actually take very long, which I was pretty happy about.

The Raven plot was pretty interesting. You can see a lot of details in the acceleration plot post-apogee.

 

[KenRico]

...

Meanwhile my Madcow Black Friday shipment showed up with the pair of 3"/38mm Aerobees and the 29mm Bomarc. Oh Yeah!!

<slow clap, growing louder & faster> bomarc, bomarc, bomarc boMARC, BOMARC BOMARC !!!!!!!!

Kenny

 

[caveduck]

Got the confirmation emails - the Beeline 900 MHz GPS tracker system is on its way! Greg noted on the order that he thinks the 900 system will work fine for the expected altitudes. Just in case I looked around and found some 3rd party antennas that could be put on the RX to get an extra 6-9 dB of gain.

Meanwhile I'm ordering some extra bulkplates from Madcow to convert the nose cone coupler into a proper secondary av-bay.

 

[MasonH]

What is the difference between this and a Wildman Extreme?

 

[troj]

What is the difference between this and a Wildman Extreme?

Similar, but I'm sure there are some dimensional differences. If you find RockSim files for both, you should be able to match them up.

Much like the Frenzy is similar to a 4" Darkstar.

-Kevin

 

[caveduck]

I'm back at it after a nice holiday season, building a glassed Madcow BOMARC and a couple of launches.

I've been anxious to try out the BeeLine BRB900 900MHz GPS telemetry package from www.bigredbee.com for a while. The system components are shown here:

I got the transmitter option with logging and the g-switch launch detection, and the receiver LCD panel for standalone use of the receiver without needing a computer to read out the position. The very small circuit board is the USB data interface/charging adapter for the TX. BigRedBee furnishes one of the needed mini-USB (*not* micro USB, standard phone chargers don't fit) cables. If you wanted to charge your TX and RX at the same time you'd need to go pick up another cable.

There was no paper or electronic documentation media included in the package; you need to go grab it from the BigRedBee website, which is fine with me (I keep a completely paperless office at work these days). Armed with the PDF I can put it on whatever device I use in the field; an Android phone and/or tablet.


The only thing you really have to know to use the unit right out of the box is that you need to plug the short power cable hanging off the TX onto the matching header on the TX board to *operate* or *program* the TX, or plug it into the charging adapter to *charge* the TX battery. It will not operate from the charging circuit. The documentation makes this clear but it's worth mentioning here.

The red wire on the right side of the next picture is the antenna. For flight it will be inserted into a rigid tube (an Aerotech ignitor shipping tube ought to be about right) to keep it straight during acceleration.

With the expected battery life being a couple of days, the recommended installation method is just to wrap up the TX in bubble-wrap and stuff it in a (non-metallic, non-carbon-fiber) nose cone. You can happily fire up the TX the morning of flight, install in the nose and just leave it on. In reality you only need the battery to last until the rocket lands unless the local rugrats run away with it and you need to hunt them down :no:

The system has quite a few programmable parameters that let you change the NMEA sentences that are sent down, increase/decrease the reporting frequency, etc. This will enable telemetry of velocities as well as position during flight if you want, at the cost of reduced battery life. However for basic post-flight location, the factory defaults should work fine.

The next photo shows the RX display while the TX has not yet acquired a position fix. In the lower right corner it reports the TX battery voltage and number of satellites acquired so you can tell how it's doing. Mine took about 5 mins to start reporting position, and about 10 mins to start reporting a fairly good position (HDOP down to 120-130), despite having been mailed down from OR to CA. Subsequent startups should be faster since it will now have current ephemeris/almanac data for this location.

 

[caveduck]

(Post continued since I have one too many photos for a single post)

Here's a pic of the RX display when the receiver has a fix and is reporting. Note that the lat/lon readouts are formatted as dddmm.mmmm - the most important thing you need to know when poking the position into your handheld GPS that you're going to use to locate the rocket. Here I blanked the most significant digit of the minutes in Photoshop as a minor privacy measure. The display looks like it will do pretty well in sunlight, though that's a bit hard to test during late-night rocketry prime time.

I turned the transmitter off by unplugging the battery from the TX, and the receiver LCD continued to display the last received position as you would hope.

BTW there is a feature to enable the TX to have its power controlled by a jumper/switch, but unless a special reason shows up I doubt I'll bother to use that since the battery life is so good.

To check the position it reported, I typed in the position to Google Maps and it geolocated to my exact address.

The last photo here shows the TX charging configuration with the battery power cord plugged into the charging/data adapter.



So far I'm really pleased with the BRB900. It seems to be very capable and does the basics with next to no user intervention, and the size and weight are very reasonable for HPR.

There are few more things to do yet before I start launching it -

• Ground range test
• TX battery life test
• Check out the TX programming interface and RX data streaming features, though I won't use them right away
• Pick and test an Android app to use on my phone to navigate to the unit in the field, probably Critter or Trimble