L3 Build - Harakiri

[degreaser]

Introduction.

The TAPs:

After having received my L1/L2 at Mudroc 2011, I approached Tony Alcocer
(TFish) at Aeronaut and requested that he serve as one of my L3 TAPs. He
graciously accepted and suggested Richard Hagen as the second. However,
securing Richard's assistance was slightly more challenging. Circa
January 2012, I made the comment on Aeropac's email list that it would be
nice if Aeropac had the ability to accept PayPal payments for membership
dues. Unbeknownst to me at the time was that a request made on this list
was taken by Richard as to one having volunteered to implement said request.
One thing lead to another and I accepted this task in exchange for Richard
being TAP #2. Conclusion: Tony's easy, Richard must be bribed. Note for
future reference.

All kidding aside, I would like to offer my most sincere thanks to these
gentlemen for any and all assistance and mentoring they may provide, past,
present and future.

The Design:

Over the intervening year since Aeronaut 2011, I contemplated several
different designs for an L3 rocket. Tony favors relatively small, minimum
diameter, high performance rockets. While that design philosophy is of
interest, I have little to no experience to date with minimum diameter
construction techniques. I also considered fabricating a 1/3 scale model of an
Honest John for use as a certification rocket. Construction of a 1/3 scale
model would be preceded by a 1/4 scale model to gain the necessary experience
in nose cone fabrication. A great deal of time was spent vacillating between
big, fat and slow vs. small and fast.


While preparing to depart XPRS 2012, I spoke briefly with Tony regarding
my L3 plans. I was discussing my tendencies to build heavy and my
frustration with building booster sections which have limited the rockets
to 4 grain cases. What I got out of the conversation was that an L3 build
was no place to begin to optimize building techniques for weight reduction
purposes. To put it simply, I would go with the same sort of techniques
which had proved successful in both my L1 and L2 builds. At this point,
the overall design of my L3 build became apparent. While the 1/3 scale
model HoJo likely won't be constructed, the 1/4 scale model is in the
works, albeit on the back burner at the moment.

I decided on a 5" airframe, 4FNC. This choice would allow the use of
proven TTW construction techniques with 98mm motor casings. I chose
to design for a 6XL grain case which would allow for the use of N and
baby O3400 motors post-certification and avoid the 4 grain case frustration
experienced with several of my earlier builds. Performance Rocketry 5"
components were chosen as they are readily available.

Two design considerations are paramount; safety and survivability.
These criteria must be kept at the forefront throughout the design
and build process. A related goal of this build is to begin to
experiment with T2T techniques and to utilize carbon fiber reinforcement
during the fin can construction. Another is to experiment with cleaner,
more modular avbay designs.

Onwards.

 

[degreaser]

The introduction in post 1 was excerpted from the design pdf which will be submitted to my TAPs as part of the documentation process. It has not yet been completed and is somewhat out of date. I'll post it once the document has been updated. I'll post one more excerpt for now.


Specifications

Material: G10/G12 Fiberglass, 6061 Al Bulkheads
Length: 124"
Diameter: 5.15"
Span Diameter: 17.15"
Fin Count: 4
GVW: 283oz Nominal
Vne: Mach 2.0
Motor Mount: 6 grain 98mm
Retainer: Aerotech Flange Mount
Fin Can Construction: TTW, T2T Carbon Fiber
Recovery: Redundant Dual Deploy
Avionics: Telemetrum
Avionics: Raven 3
Avionics: BRBGPS70
Avionics: Icarus 1 (under development)
Onboard Video: GoPro Hero 1

As this is a certification rocket, this rocket will be overbuilt. It will be fairly heavy but I'm more concerned about having it come down in one piece regardless of the motor I use rather than squeezing out performance and altitude from the design. I'll save high performance optimizations for later builds.

 

[UPscaler]

awesome! Will b watching this thread!

Also, if I may make a suggestion (unless you already have the GoPro)...The Contour Roam 2 shoots a higher frame rate at 1080p than the GoPro Heros 1,2 and 3. (3 being the white edition, of course. The black edition is awesome, that's a bit too much for a rocket camera to me...) :smile:




Braden

 

[degreaser]

The first subassembly to be constructed is the fincan.


The stock centering rings were a variety of thicknesses. I decided to epoxy several together for mounting the motor retainer and to provide additional support for the fins.

To help ensure that the rings are epoxied concentrically, I used a rotary table to drill alignment holes. The centering ring in the second image is being reduced in outside diameter to fit inside a coupler section at the top of the fin can. All of the other centering rings fit inside the lower airframe.

 

[degreaser]

The lower centering ring consists of a sandwich of three centering rings for a thickness of .5". The stack was drilled and tapped for 8-32 SHCS as supplied by Aeropack. As the holes were fairly close to the OD of the centering ring, I supported the circumference of the ring with a hose clamp. The clamp helps to prevent fracturing of the ring when drilled and tapped.

 

[degreaser]

Next comes the actual mounting of the rings onto the 4" (98mm) motor mount tube. The tube is 48" in length to allow the use of 6XL cases. The lower centering ring was epoxied in place using a bead of JBWeld. The remaining rings were tacked with Devcon 30 minute epoxy.

 

[degreaser]

Next comes fin beveling. The big problem with beveling is how to do it. I've seen a variety of techniques ranging from hand sanding, disc and rotary sanders etc. I chose to fabricate a fairly large sine plate and use my mill to do the beveling.

A sine plate is nothing more than a flat plate with grooves machined into the lower surface. The distance between the two grooves is "h". Round bars are set into the grooves and one end is raised by a predetermined amount "y". The angle is determined by the formula arcsin(y/h). Hence the term "sine plate".

Commercial sine plates are quite expensive. This one was fabricated from a 22" length of 1"x6" 6061 Al bar stock. I used a 1/2" carbide 90 degree V-groove router bit to mill the slot. The plate was drilled and tapped 3/8-16 in a 2x2 grid. The tapped holes are used to secure the plate to the table and the workpiece to the plate.

 

[rocketkyle]

Nice design! Looks very similar to my L3 rocket. https://www.freewebs.com/rocketkyle1/l3rocket.htm
Good luck, should fly high and fast.

 

[degreaser]

Now for a test of the fin beveling.

As I don't have facilities for accurately cutting large sheets of G10 material, I had PML fabricate the fin blanks for this rocket. I asked them to include a few scraps of material for test purposes. The material is approximately .206" thick.

I supported the sine plate with 3/4" drill rod and raised the back edge 1" with a couple of 123 blocks. This corresponds to a bevel angle of arcsine(1/5) or 11.5 degrees. Using a 1/2" carbide two-flute endmill, I set the depth of cut to .060". This depth will not result in a knife edge; I don't want a knife edge as I want to minimize the risk of damage should the rocket be drug across the playa for some distance at Black Rock. I used a mist of coolant to keep the fiberglass dust down to a minimum.

 

[degreaser]

Time for sweaty palms. The beveling of the fins.

I milled slots in the root edges of the fins to clear the two centering rings in the middle of the root spans. Then I beveled the other three edges.

 

[degreaser]

Now that I have a motor tube with centering rings and beveled fins ready to mount, the question is how to accurately mount the fins.

For my L1/L2 builds I fabricated a small alignment fixture which was secured in the spindle of the mill. The XY milling table was adjusted to put the center of the fin over the center of the motor tube. The motor tube was rotated using a rotary table. The result was a reasonably accurate, rigid structure that would hold the fin in place while the epoxy bead cured.

Unfortunately, the fixture was a little small for this rocket. That means a new fixture was needed.

I modeled the fixture in Rhino and then fabricated the fixture. It's 22" in length and can be adjusted to different lengths fairly easily. It is machined from 6061 bar stock. In this case, the fixture is attached to the spindle mill with a length of 3/4" drill rod via the large bracket on the side of the fixture.

 

[degreaser]

And now the actual fin mounting and alignment.

My rotary table is 6" in diameter. This rocket has a fin span diameter of better than 17". Obviously the rotary table must be raised a fair bit for the fins to clear the table. I grabbed just about every squarish bit of material I could find and managed to raise the table sufficiently to clear the fins. I also needed to extend the table to support the end of the motor tube. I used some 3/4x6" Al stock to do the extension.

Each fin was attached with a bead of Devcon and allowed to cure 24 hours before the next fin was attached.

 

[degreaser]

The interior fillets are next. The fillets are formed with a 50/50 mixture of Devcon with milled fiberglass. I smoothed the fillets with a length of 3/4" delrin rod which had been rounded at the end.

My first set of fillets became contaminated with denatured alcohol and wouldn't cure properly. I was forced to chisel and scrape the entire set of fillets out and re-prep the surfaces. A total PITA. You should try it sometime.... (once).

I will eventually attach a complete set of images of the fillets for documentation purposes.

 

[degreaser]

Next comes the slotting of the lower airframe. This is the second set of sweaty palms.

First, I had to make some mandrels. I cut these from 1" plywood using a Jasper jig and then turned them to a snug fit on my lathe. I made two sets of 6 mandrels; one set for the airframe tube diameter and the other set for the coupler tube diameter.

As you can see in the second image, mounting a 48" length of airframe tube on a 30" table can be challenging. I used the sine plate to secure the rotary table and the length of 3/4x6" Al barstock to secure the far end of the tube. The mandrels are mounted on 1/2" allthread.

The slot was cut with a 15/64" carbide end mill.

 

[degreaser]

And the test fit! It finally looks like a rocket!

 

[degreaser]

I pulled the test fit apart for further work on the internal fillets.

I've attached the first shot, more to follow. The fillets are actually done, I've just not taken photos of them yet. You'll note the lengths of 10-32 stainless allthread. There are four of these and will help to secure the upper bulkplate. The allthread and nuts are secured with red loctite and then the fillets formed with a 50/50 mix. As indicated earlier, this is a cert rocket and will be over built. The last thing I want to see is the bulkplate being pulled out during deployment.

 

[degreaser]

Before I can permanently mate the fincan with the lower airframe, I need to fabricate some rail button mounts. Drilling and tapping into a .0625" thick fiberglass wall might hold a rail button but I'm not going to trust it. Instead, I'm going to mill some mounts from 6061 stock, then drill and tap to 1/4-20. Two mounts must be fabricated with slightly different radii. The lower mount must match the ID of the airframe tube and the upper mount the ID of the coupler tube. Both must also match the OD of the motor mount tube.

The first image shows a Rhino model of two possible designs. I decided on the square design as it's easier to mount on the rotary table using things that I have on hand. It will be fabricated using 1" square stock.

The second image shows a rather useful widget - it's called a coaxial indicator and is used to align the center of the spindle with a round hole, in this case the bore of the rotary table. Once the center of the table is indicated, one moves the X (or Y) axis out the desired amount to form a given radius on the cut.

That's all for now... TO BE CONTINUED....

 

[degreaser]

awesome! Will b watching this thread!

Also, if I may make a suggestion (unless you already have the GoPro)...The Contour Roam 2 shoots a higher frame rate at 1080p than the GoPro Heros 1,2 and 3. (3 being the white edition, of course. The black edition is awesome, that's a bit too much for a rocket camera to me...) :smile:




Braden

I already have the GoPro Hero. I'll have to take a look at the Roam though.

 

[degreaser]

Nice design! Looks very similar to my L3 rocket. https://www.freewebs.com/rocketkyle1/l3rocket.htm
Good luck, should fly high and fast.

Thanks! 4FNC's tend to look similar

 

[MasonH]

AWESOME build.

 

[dixontj93060]

Man, amazing tools and jigs!!

 

[Eric1]

Nice work Degreaser.

 

[degreaser]

Thanks guys!

 

[degreaser]

Fillets, fillets and more fillets. 5000K lights!

 

[CarVac]

Fillets, fillets and more fillets. Pardon the color balance. Warm fluorescents vs cool flash.

You're pardoned for at least knowing that white balance IS an issue and what's causing it. (I wouldn't have the gel to tint the flash either, haha)

Anyway, awesome work!

 

[degreaser]

And even more fillets!

 

[degreaser]

You're pardoned for at least knowing that white balance IS an issue and what's causing it. (I wouldn't have the gel to tint the flash either, haha)

Anyway, awesome work!

Thanks. White balance drives me nuts. The fluorescents over the workbench are warm, the fluorescents on the ceiling are cool as is the flash. I'm always fighting one or the other.

Ok, I'll break down and get 5000K bulbs all around. I've been putting it off waiting for the existing bulbs to give out. It could be awhile.

 

[degreaser]

Ok, fixed my lighting problem. I replaced all the fluorescent tubes with 5000K natural light tubes. Good riddance to blue highlights and orange shadows. On to the railbutton mount machining...

The first image shows the setup. I'm using a 1/4" two-flute carbide endmill with a squirt of wd40 for lube. The large radius is cut first as shown in the second image. There was a bit of chatter but I'm not too concerned as I plan to rough up the surfaces with 80 grit prior to epoxying them to the tubes. The third image shows the second cut nearly completed. The astute reader will note the tooling change between the second and third images. I used an er20 collet and holder to gain a bit of z-axis clearance as the spindle was hitting the step block at the rear of the table. The completed pair is shown in the fourth image.

The last image shows that it clears a straightedge between two of the centering rings. I cut the radii .005" off on each side for a total of around .010". This is to allow for the epoxy between the mount and the tubes.

Now, should I drill and then epoxy, or epoxy and then drill... There are pros and cons to either.

 

[cbrarick]

Great build thread and you have even better equipment! I'm jealous!

One comment, you don't need to make your motor tube as long as the longest motor you want to use, just the fin can (plus the length your ebay sticks into it)

rick

 

[patelldp]

My OCD approves!

Seriously, great job! I love the lines of those fins.