To put it simply, there are a lot of aerodynamic surfaces on my New Glenn model. The most important are the main fins, which are located at the aft end of the rocket. These fins are 1/4″ thick clear acrylic fins (currently still in their protective blue plastic cling wrap). They are of a through-wall design, so they will be epoxied to the motor mount tube as well as to the inside of the aft module. A small fillet of epoxy will also be used on the outside joint. They have a root chord of approximately 10.8 inches, a tip chord of 5 inches, and a height of 8 inches (measured from the 6 inch primary airframe). The root chord is contoured to match the profile of the rocket as it transitions from the 6 inch primary airframe to the 7.7 inch aft module. While alignment is handled primarily by the slotting, the fins also utilize locking tabs, which fit into laser cut notches in the middle and aft centering rings. The primary function of these tabs is to keep the fins square with the airframe during construction, but do provide an extra level of assurance for minimizing fin cant.
The second set of fins are the forward canards. Located 29.5 inches from the nose end of the rocket, these fins actually provide a destabilizing force. For this reason, the rocket will have to fly with at least 1 kg of payload. Each of the four canards has a root chord of 4.3 inches, a tip chord of 1.4 inches, and a height of approximately 1.8 inches to maintain the scale of the rest of the rocket. These are made from 1/4 ” thick plywood and are painted white and blue. Unlike the main fins, the canards are surface mounted so as not to interfere with the main parachute, to whose bay they are mounted. To provide additional attachment strength, each canard will be held by two pair of 3D printed PETG brackets with 10-24 binding posts. These brackets will be epoxied onto the airframe and provide approximately 1 square inch of bonding surface per bracket pair. Due to launch rail considerations, these canards will be mounted 45 degrees off their intended position, aligning with the main fins.
The final set of fins are the strakes. These two fins are also made from 1/4″ thick plywood and painted white and blue. Like the main fins, they are contoured to match the profile of the transition in the aft module. They have a root chord of approximately 14.9 inches, a tip chord of approximately 2.8 inches and a height of approximately 2.6 inches to maintain scale. Like the canards, the strakes are surface mounted, but this is because they were not originally considered for inclusion in the model, so the appropriate accommodations were not provided. To provide additional attachment strength, each strake will be held by two pair of 3D printed PETG brackets (same as used for the canards) with 10-24 binding posts. The strakes, each 180 degrees apart, will be mounted 45 degrees off of the main fins and 90 degrees off from the launch rail.
Aerodynamics were not a major consideration (or rather, more drag was considered better to lower the max speed and apogee for a certification flight), so all fins will keep their square leading and trailing edge profiles. This also has the benefit of decreasing the work required on the fins and maintaining the same standard of quality for each fin.
As previously mentioned, the payload bay and aft section airframes of my New Glenn scale model are not of a standard size, so I will have to make those tube sections myself. In order to make the airframe tubes, I will also have to make the mandrels, which are to be the subject of this post. I previously printed a few mandrel negatives so I could cast the mandrels with polyester resin. I think I made an error printing the aft mandrel negatives, so I will have to re-evaluate and try again. Each negative is 3 inches tall (to save on printer material), so mandrels will be cast over multiple sessions. I bought a couple of 2 foot segments of 4-inch PVC piping to reduce the amount of resin I would need to cast. I did a sample pour with the forward mandrel to prove my concept and noticed a few things. First, it was difficult to maintain a good seal at the bottom. This could probably be solved by planting the PVC and mandrel negatives in a base of clay or something similar. I also imagine this would only be an issue in the first pour. Second, the fumes from the polyester resin were overwhelming. I think the only properly-ventilated place to do this task is outside. The fumes also stick around a bit after the resin has cured, so I’m not a big fan of this method. The results of my first pour were pretty encouraging, though. I only used a few ounces of resin (so as to avoid waste from rookie mistakes), which is why the segment in the photos is so short.
As you can see, the alignment between the mandrel and the nosecone (which will fit into the tube made by this mandrel) is pretty good, so I am encouraged by the results. Following the casting of this short segment of mandrel, I purchased some air-dry modeling clay, which I think will actually be much faster to cast and have no fumes. To increase the hardness of this clay mandrel, I will glaze the clay mandrel with a thin layer of polyester resin (outside), which should also give a a few thou of tolerance between the manufactured airframe tube and the parts that mate into it.
I made the decision to 3D print the transition sections using white ABS. My printer, a FlashForge Creator Pro, is a pretty nice unit with a fairly large build area–approximately 8.9 inches by 5.7 inches on the base, and 5.9 inches vertically–but this is also a really big rocket. Neither of the transitions could fit on the print bed, and even if they could, there still would not be enough room in the Z direction. My solution to this problem was to divide each transition into four segments. Additionally, the nose cone, while it could fit on the base, was too tall for my printer, so it too got sliced in half.
The immediate question that arises after splitting up a large print is how to join the pieces. Joining with an adhesive (say, epoxy), plastic welding, and melting the mating surfaces together with acetone were all viable options, but left some risk of the joint failing catastrophically, especially given that being such a large print, there was an increased probability for a wavy surface. (I’m still working on dialing in the bed level and print settings, but generally the mates are pretty good.) My solution was to employ one of these methods (leaning towards acetone melting), but also to design in alignment pins, which would relieve some mechanical strain from warping and other deformations. This would also allow for the optimal alignment of the parts for better surface continuity at the segment interface. Below are the pre-joined nose cone, forward transition, and aft transition.
After printing, I bored out the holes a bit with a 9/64 drill bit so the 1/8 pins would slide in. The pins (1.5 inch long steel) are secured into one of the segments with a drop of CA glue and then the opposite (horizontally) segment is mated to it. I have not joined these segments together yet; that will occur once I verify they fit snugly into their respective airframe tubes. For the transition units, the top and bottom halves will be mated such that the horizontal interface planes of each are perpendicular. This should theoretically increase the strength (or, rather, decrease the weakness) of each transition unit. I noticed when trying to join the top and bottom halves of the forward transition that the pins did not align and the diameters at the mating plane may have been slightly different. I figure I made an error when breaking up the forward transition, so this will have to be re-printed. Below are the aft transition, forward transition (the horizontal mates looked nice), and the nose cone when joined together.
I noticed that there is a little bit of sag in the prints, so I will have to do a bit of cosmetic work (sand and coat with epoxy) to the top-bottom mating interfaces. I anticipated this might happen, so that dictated the decision to print all components with the top-bottom interface at the bottom of the print. This meant that I would not compromise the circularity of the transition shoulders or the effective diameter, since the mate into the airframe tube is a far more important mate (and more difficult to repair) than the mate between each segment.
I’m working on a number of exciting new rockets to bring to NSL this year, but the one I’m most excited about is the scale model of Blue Origin’s yet-to-be-built New Glenn rocket, since this would be my Level 3 cert. The New Glenn rocket is going to be pretty big , so I scaled it down from the photos out there such that the majority of the airframe would be the size of a 3″ PML tube. After looking at it for a while, I figured I might as well double the size and stick an M in the back of it to make it an L3 rocket, so most of it is the size of a 6″ PML tube. At just over 77″ in length, it’s not the tallest rocket on the block, but it still should be pretty awesome. I’ve baselined the rocket to use an AT M1350 motor, but the motor mount is a 98mm mount, so I can add more power later if I want (and have the $$$). Projected altitude is just over a mile because I believe in low and slow cert attempts.
So what’s the construction plan? The only thing that fits well into a standard tube size is the 6″ airframe and the 98mm motor mount, which will both be PML phenolic tubes (no glass). Unfortunately, it seems that Blue Origin did not have rocketry component scales in mind when designing this rocket, so I’m getting a bit more creative with the payload fairing and the aft section. In both cases, I am attempting to cast a polyester resin mandrel (4.9″ for the payload fairing and 7.1″ for the aft section), from which I will then use to create a short section of fiberglass airframe. If the cast mandrels don’t work out, I may have to 3D print the mandrels (which I don’t want to do) or find a way to turn a mandrel (I don’t have a lathe). A lot of other things are actually already 3D printed–the nose cone and both transitions, as well as some internals. The canards and strakes will both be 1/4″ laser cut plywood, and I’ve added four additional acrylic fins (1/4″ laser cut as well). Most of the centering rings and bulkheads are stock from PML, but I also got a couple of custom ones laser cut to TTW fin locking and for the non-standard diameter airframe sections.
Recovery will be standard dual deploy (36″/96″) with redundant altimeters. I’ll probably fly with my Raven as a primary and possibly use the Eggtimer I recently built as secondary. (I could also use the RRC3, HiAlt45k, or Eggtimer Quark.) I will likely leave my Eggfinder off, as I don’t anticipate too many issues finding a low-flying rocket in the desert, but will play it by ear on launch day. I will also fly with a 2kg dead weight payload for added stability and decreased altitude.
The final step in building my Level 3 rocket was installing everything into the airframe. The airframe is a cardboard tube (no fiberglass or carbon fiber reinforcement). The decision was made to only use cardboard since the vehicle would remain subsonic, and no composite reinforcement would keep the weight down and weight distribution favorable. The airframe consumed two LOC 5.38″ (OD 5.5″) cardboard tubes. A whole tube went to the booster section. The second tube went to the two parachute bays and two switch bands. Three fin slots were cut into the booster airframe tube (painstakingly and with an Exacto knife–blood, sweat, and tears all contributed to the building of this fantastic machine!) from the aft of the airframe to the top of the fin slot.
After a quick dry fit, I drilled holes for the two 1515 rail buttons (courtesy of Dog House Rocketry). The rail buttons are held in place by two expanding rubber well nuts.
The aft rail button is 3 inches from the aft of the airframe. The forward rail button is 5 inches ahead of the CP, or 35 inches from the aft of the airframe, or 6 inches behind the nominal CG. (The nominal modeled CG was at the edge of the booster coupler and there were also spatial constraints imposed by the forward centering ring.)
In honor of my alma mater Georgia Tech, the rocket got a white and gold paint job. Now it’s time for the Yellow Jacket to take to the skies at AirFest 22 in Argonia, KS! I hope to do a first flight on an L motor (possibly a CTI L1395 I got from Chris’ Rocket Supplies).
My Level 3 rocket has three segments (a booster and two parachute bays), so it also needs two couplers to mate all the segments. For this, I used a pair of standard LOC 5.38″ cardboard couplers, which are 11 inches in length. I knew in advance that one of these would be purposed as an avionics bay (and I wanted both to look roughly the same), so I cut off a couple of 1 inch bands from the tube (a LOC 5.38″ cardboard airframe tube) from which I cut the parachute bays to act as switch bands. I glued these bands onto the coupler 5 inches from either end with wood glue.
I have a fairly small car, so I designed this rocket to be disassembled into multiple fairly small pieces. However, this means that as assembled, this rocket may require extra precautions to ensure everything stays together during all phases of flight. For this, the avionics bay coupler is held to the main parachute bay (forward) by twelve pop shank rivets in two offset patterns of six rivets, and to the drogue parachute bay (aft) by the same. (The planes of separation are at the forward end of the main parachute bay and the aft end of the drogue bay. The booster coupler is riveted with the same arrangement to the booster segment. The booster coupler and drogue bay are constrained by up to six #2 shear pins (nominally three are used). Since the airframe and coupler components are primarily cardboard, special consideration is given to preserve the life of these components. In particular, the couplers are thinner than the airframe. PLA backing rings were 3-D printed for the interior of the couplers for the riveted and shear pin joints and super glued in place. These should minimize the effect of dragging rivets and shear pins through the cardboard couplers. Additionally, the number and alignment of rivets was chosen to distribute the shearing load through the airframe.
I had originally planned to have two key switches to arm the altimeters, so I painstakingly drilled the two holes for those as well as two holes for LED indicators. Unfortunately, I discovered that the solder joints on the key switches I bought were a bit brittle and (more importantly) that the actual switch seating mechanism got busted after some number of turns (it appeared due to some seal or rubber component in the assembly unseating and jamming elsewhere in the mechanism. This inspired me to switch (pun intended) to a simpler, more elegant solution with PCB screw switches. For this, I 3-D printed a switch housing for the screw switches for easy accessibility as well as holes to mount indicator LED lights. Finally, I printed a switch housing cover, which is friction fit into the housing. The switch housing is the exact width of the switch band and a little less than a quarter the circumference of the switch band.
In order to get the indicator LEDs to function properly with any altimeter, a simple logic circuit involving a 5V regulator (I blew out too many transistors) and a resistor was created. Two of these simple circuits fit onto a small proto board. The proto board was fitted into a custom 3-D printed bracket, which was then epoxied just above the aft plastic backing ring on the avionics bay coupler. V+, V-, and ground wires were routed to the forward end of the coupler where they could be plugged into a terminal block on the each altimeter canister. V+ and V-, and ground wires were also routed to the corresponding indicator LEDs and screw switches.
The avionics bay coupler is capped by two 1/2″ plywood bulkheads. Each of these bulkheads has two holes for the altimeter canisters, which are housed in standard 38mm tubing. The canisters are retained with laser cut 1/4″ plywood brackets and four #6-32 stainless steel screws (and corresponding tee nuts) per bracket. Each bulkhead also features a 2″ steel U-bolt with 1/4-20 threads. The MAIN (forward) bulkhead also includes a terminal block with indicators for which wire to plug in where, whereas the DROGUE (aft) bulkhead contains no such feature. Ejection canisters are integrated into the altimeter canisters.
To complete the avionics bay coupler, I drilled four 1/4″ vent holes in the switch band in 90 degree intervals. I super glued in 5mm LED holders into these holes to preserve the integrity of the holes and artificially remove the burr from the cardboard.
The booster coupler features a 1/2″ plywood forward bulkhead with a 2″ steel U-bolt that ties into the drogue parachute system. Because the aft part of the coupler carries no loads, there is only a 1/4″ plywood aft centering ring less than 0.5 inches wide to add rigidity to the aft coupler.