Quantcast

Eyelash: an altitude seeking minimum diameter I class scratch build

The Rocketry Forum

Help Support The Rocketry Forum:

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Hi. I'm Kelly.

This is Eyelash:
IMG_7975crop.jpg IMG_8476crop.jpg img_8326_crop.jpg

Eyelash is a scratch built minimum diameter rocket with the goal of flying high and testing some new concepts while developing skills and having fun. Most of the rocket was constructed out of hand laid fiberglass so it wan't a minimum effort build.

I thought I would start an Eyelash post to provide details about the build and building process so you can learn from what I did and teach me what I should have done. I'll try to post construction details of most components over the next week or two.
 
Last edited:

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Design Process

Designing Eyelash involved iterating among three things: simulations, rocket notes in Excel, and making the rocket components. Plans evolved as components were created and their final dimensions and masses became clear. At the end all match and simulations can be compared with real flight data.

I used both RockSim and RASAero II for simulations. I like RockSim for the ability to look at how the parts fits together. The component modeling allows a decent estimate of the center of mass that cannot be generated in RASAero. RockSim also does a reasonable job of predicting altitude for subsonic flights. I consider OpenRocket to be interchangeable with RockSim. In my experience RASAero is more accurate with supersonic simulations predictions and CP estimates. The settings “Rodgers Modified Barrowman” equations, all turbulent flow, and "smooth paint" do a pretty good job for what I build.

For an altitude seeking rocket I made the whole design around a specific motor selected from the CAR/NAR/TRA certified motor list. I chose the CTI I216W because it is a full I and only 38 mm diameter. The drawback is high thrust, but there are tradeoffs and my ideal motor doesn’t exist commercially. Last year I made a high flying rocket around the I65W because I love the nearly 9 second burn, however it has a 54 mm drag penalty. I wish a 38 mm, long burn, full I were available. Hint hint: Aerotech, CTI, Loki.

Since I started flying rockets in 2016 I have kept detailed notes in Excel spreadsheets on each rocket build. This information allows me to come up with decent initial estimates of the properties for the components I will build. I use one tab to collect notes about how I made each component: which fabrics, epoxies, details, processes etc. On another tab I store the length, width, thickness, diameter, location, mass, and other details on each rocket component. This is the budget I work from for mass and size. Usually components are broken down into groups with a mass sub budget. For example with the av bay I will have the masses each little piece all measured, listed and tallied: sled, computer, battery, ‘coffin’ for battery, wires, screw switch, screws for mounting the computer. For fins I record the initial fin weight, mass added during bonding and filleting, and tip-to-top. When I make a new rocket, I can look back to the details of previous rockets and make a pretty good estimates in terms of size, stiffness, and mass for initial design and modeling.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Airframe Tubing

I started learning to make fiberglass composites from posts on this forum and movies on the internet. Joey provides a few references to some of these sources, and a lot of his own build techniques, in his Rocket Ship 27 build. Thanks to all who have contributed. Sharing your hard earned lessons has really helped myself and others progress.

The airframe in Eyelash is constructed out of 3 layers of 3.7 oz S glass and a top layer of 1.43 oz regular fiberglass. The top layer provides a smooth finish and serves as sacrificial sanding layer to protect the structural layers. I used Aeropoxy PR2032 resin with PH3663 hardener.

fabric.JPG

S glass is more difficult to cut than regular fiberglass cloth. I found a pair of poultry shears in the kitchen that do a pretty good job. They have serrations that grip the S glass and keep it from slipping while cutting. This helps a lot on small or narrow pieces.

shears.JPG
A 1.5 inch OD tube purchased from Online Metals served as the mandrel. A pipe or tube from a standard supplier with a little sanding works great as a mandrel. Even though case sizes are nominally specified in millimeters, their true dimensions match inch fractions. Here are CTI pro 38 case dimensions. OD = 1.500"

For release, the mandrel tube was wrapped with 1.5 layers of 2 mil clear mylar. Sometimes I use 2 or 3 layers to get the tubing ID that I want. 1.5 layers is about the minimum and is all I needed for this rocket. Never tape the mylar to the mandrel in the middle. You will never get the completed tube off the mandrel without destroying it. I do tape the outer edges of the mylar to the mandrel Outer edge tape helps keep the mylar from sliding and achieve a tighter wrap. I use a strip of double sided scotch tape the entire width of the mylar wrap to stick the trailing edge to itself. This helps keep the edge of the mylar down flat and epoxy from wicking in.

I degassed my epoxy in a small vacuum chamber I made out of ABS with a rubber gasket and a piece of polycarbonate for a window. I used a ¼” NPT tap directly into the ABS as a vacuum connection. Cut ABS has a porous surface that doesn’t seal well. To make a smoother rim I smeared ABS cement on it and let it harden against a sheet of glass coated with PVA mold release. This is about $10 worth of materials, assembles quickly, and is a size for degassing the quantity epoxy I typically mix. In the case of a bad mess it could replaced rather than cleaned. TAP plastics sells polycarbonate scraps by the pound. Veneer supplies is a good source for vacuum fittings such as the quick disconnect I used.

degass.JPG

I have started testing a longer chamber for degassing completed tubes, but I haven’t decided if it adds benefit. This rocket tube was not degassed after wrapping.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
I begin wrapping the tube by painting a coat of epoxy on the mandrel. Then I tack on the edge of the glass cloth and add epoxy with a disposable brush while winding more cloth on. I wind the mylar, all layers of fiberglass, and peel ply in the same direction. This allows gentle tugging to tighten the entire bundle and eliminate wrinkles and bubbles. I like the unwrapped material to hang in front. I usually brush on epoxy starting in the middle of the tube and working towards an outer edge and downward. This helps prevent bubbles and wrinkles and get a tight bundle. You want the cloth completely saturated without bubbles and gaps but not over saturated and dripping. Tapping in epoxy can help fill voids. There are a number of videos online that show this process better than I do it.

wrapping.JPG

While building up the layers I occasionally go over the surface with a heat gun. Be gentle and don’t over cook it. With the right amount of heat and timing the small bubbles that cause pinholes pop and the epoxy loses viscosity and flows well. I find that the Aeropoxy responds really well to warming without losing much working time. When it’s too cold, like the ~60 degree room temperature in my basement, Aeropoxy can be a little stiff and clog the brush. I used to use West System 105/206B which applies easily without heating but gets viscous quickly when heated. My main reason to switch was the higher glass transition temperature of Aeropoxy.

I finish with two complete wraps of peel ply. I try to make these tight to squeeze as much epoxy out of the tube as possible. I add just enough epoxy to fully wet out both layers. Try to leave the tube smooth, no wrinkles and no visible big bubbles.

peel ply.JPG

I cured this tube with the mandrel vertical. I would rather have the epoxy flow towards one end of the tube than an edge. For the first few hours I flipped it end for end about every half hour.

Here is the tube post cure and before any sanding.

raw tubing.JPG

I made a fixture to rough sand the outside of the tube. This sanding method takes off proud areas like the edge seam with minimal impact to the rest of the tube. A more uniform cylinder results. The fixture helps reduce clogging by allowing sanding debris to fall out. I start with something rough like 100 grit and then go to 150 grit. Use a low RPM to keep the heating low and reduce the risk of injury. Sand sharp edges off the fixture. Wear a proper dust mask.

honing.JPG drum.JPG

The mandrel of a 1.5” sanding drum was used to drive this tube during the coarse sand.

Shorter tubes can be sanded in the drill press. I have used a hand drill with longer ones. Sometimes I leave the mandrel in the tube for the coarse sanding. This helps add rigidity to really thin tubes and gives me more diameter options for the sanding drum mandrels.

After the rough sand I slather a thin layer of laminating epoxy on the tube which helps fill in scratches and pinholes. The rest of the sanding is by hand.

The body tube produced is 38.60 mm ID, 39.60 mm OD, and 0.102 grams/mm. Quite by accident this is the exact ID specified for Madcow 38 mm tubing. Their thin wall tubing has a 40.64 mm OD and a mass of 0.239 grams/mm.

I made tubes of a few different thicknesses and selected this one for Eyelash. I chose one that in my opinion has enough stiffness for the task. It would probably not survive coming in hot, but will allow more mass to shift to the nose. A thicker tube would also have been fine in this rocket.
 
Last edited:

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Coupler

I made the coupler from a piece of body tube that was slotted so to fit inside the original tube. The seam was glued with a single layer of 3.7 oz S glass that hardens with the coupler inside the original tubing to get the OD correct. Mylar keeps the coupler from sticking to the tube.

In the past glued the seam inside the actual airframe. I have transitioned to cutting a few ~1/4 " lengths of tubing that I space out along the coupler during gluing. End cuts off the rough tube work great.

coupler1.JPG coupler2.JPG

I like this arrangement better because I can use less mylar, maybe just a 3/4" wide strip that doesn't go all the way around the coupler. This results in a tighter fitting coupler without risking a permanent bonding of the coupler in the tube in the wrong place.

The pictures above are of a different coupler, not the one in Eyelash.

I just found an Eyelash coupler construction picture:

e coupler.JPG
 
Last edited:

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Fins

I made sheet stock for the fins out of 5 layers of 6 oz S glass, 2 layers of 3.7 oz S glass, and 2 layers of 1.43 oz regular glass. I cut the layers to different widths and align one edge so that the thickness of the finished sheet is tapered across the width. This allows the root of the fin to be thicker than the tip without sanding down a uniform sheet.

fabric.JPG layers.JPG

At one point I made a bunch of samples of different numbers layers of 1.43, 3.7, and 6 oz. I measured and weighed each sample and did curve fits to determine how much each layer contributes. I can use this to achieve the desired thicknesses across the profile. It is also useful to be able to bend the test samples and assess rigidity.

The layers hardened between two sheets of mylar, between two sheets of MDF, and were pressed with a precise 42 weight.

weight.JPG

With a quick sand on each side the fin stock came out with a flat surface and ready for next steps. The thickness varied from about 1.4 mm at the thickest edge to 1.0 mm at the thinnest edge.

stock.JPG

I use a homemade jig in the miter slot of the table saw to rough trim the fin stock and to cut out the fins. This keeps fingers away from the blade. The clamps can be screwed on in different places. I also will screw on wood strips at angles to make a fence that gets angles consistent for multiple cuts.

trim jig.JPG
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
I have recently started using a wet grinder for beveling the leading and trailing edges of small fins.

beveling.JPG fin.JPG beve;.JPG

Before gluing on the fins, I put a ring about 6 mm wide and 0.5 mm thick around the motor end of the body tube. This is just a section of body tube that I cut off. I put it on with Aeropoxy. This ring increases the OD of the back end out to the OD of the motor retaining ring and adds a little strength.

ring tail.JPG

Fins and the body tube were roughed up with 80 grit sandpaper prior to attachment. All of the shiny coating had been fully removed in the sanding process while being fabricated. I glued the fins on with G5000 Rocketpoxy and a Wildman fin jig. These fins are thinner than what it was designed for, and tapered, so I use tape and cardboard shims to make sure they are centered in the slots of the jig. The front end of the body tube is propped up to the exact height that makes the two fin jigs align on a flat surface. I usually let the first fin harden before gluing on the rest of the fins.

tack fins.JPG
 

Handeman

Well-Known Member
TRF Supporter
Joined
Jan 18, 2009
Messages
7,721
Reaction score
315
Location
Stafford, VA
Fins

I made sheet stock for the fins out of 5 layers of 6 oz S glass, 2 layers of 3.7 oz S glass, and 2 layers of 1.43 oz regular glass. I cut the layers to different widths and align one edge so that the thickness of the finished sheet is tapered across the width. This allows the root of the fin to be thicker than the tip without sanding down a uniform sheet.

The layers hardened between two sheets of mylar, between two sheets of MDF, and were pressed with a precise 42 weight.
How do you get the outer part thinner and tapered if you are pressing everything between to flat sheets of MDF? Wouldn't everything end up the thickness of the thickest part? No compression on the thinner part?

If the top is angled from the thinnest to the thickest part, how do you make sure the upper MDF isn't in the wrong position? Doesn't it end up with one flat side and one angled side?

I'm curious how you get the same angle and taper on both sides of the material.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Fillets were made of G5000 Rocketpoxy and applied between areas protected by blue tape. They were smoothed by a gloved finger in denatured alcohol. Then they were sanded so tip-to-tip would adhere.
fillet tape.JPG fillets.JPG

For tip-to-tip I applied one layer of 3.7 oz S glass that went mid way up each fin. Over the top a full layer of 1.43 oz glass was applied and it was topped off with peel ply. Aeropoxy was used.

tip to tip fabric.JPG Layer of 3.7.JPG peel ply.JPG

After hardening the surfaces were filed and sanded down to shape and smoothness.

Post sanding fins were about 1.8 mm thick at the root, 1.6 mm thick in the middle, and 1.1 mm thick at the tip.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
How do you get the outer part thinner and tapered if you are pressing everything between to flat sheets of MDF? Wouldn't everything end up the thickness of the thickest part? No compression on the thinner part?
I have made four sheets of tapered fin stock so far and they have all come out tapered, despite being smashed flat between two boards. There were not any extra shims under the top board to induce taper. They just come out that way. I think this weight really smashes the layers together and the different numbers of layers in the glass weave is enough to hold the top board tilted.

I have thought about trying vacuum bagging. I have everything I need except the bags and some kind of seal. This is working so well for me that I am not really motivated to switch.
 

Handeman

Well-Known Member
TRF Supporter
Joined
Jan 18, 2009
Messages
7,721
Reaction score
315
Location
Stafford, VA
I like that, I might have to give that a try on my next scratch build.

I kind of do a similar thing when doing tip to tip.
the first piece is 6 oz and goes from 2" from each fin edge across the tube and 2" from each fin edge on the other fin.
The next is 6 oz and goes 1" from the fin edges of both fins.
The last piece is 3 oz and cover both fins completely.

This doesn't give a lot of taper to the fins, but enough to change the thicknesses and stiffness across the fins to make them much stronger and less susceptible to flutter.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Electronics Bay

Eyelash is a zipperless head end single deploy main at apogee rocket.

sled parts.JPG top.JPG bottom.JPG closure.JPG

The sled was constructed out of 1/16" G10 and glued together with G5000 Rocketpoxy.

Most of the sled was cut out on the tablesaw with the jig in a previous post. I follow that up with a scroll saw for the fine cutouts, then rasps and files. I have an XY table I can mount on the drill press for making rows of uniformly spaced little holes. These make it easy to secure wires either by threading them through or with cable ties.

Centering rings/end caps were cut out using a fixture on a router.

router.jpg

The lower fixture is constrained to slide in the miter slot. The G10 has a 1/16" hole in the middle. This spins on a drill bit mandrel sunk into the lower board while it is cut with the router bit. Fine diameter adjustment is achieved by adjusting the gap between the fixture and a fence clamped to the router table. I have a little adjustable shim that makes it easy, but you could work with wedges and other shims like paper.

First the sled (minus screw switch) was made and the coupler was prepared with tapped mounting tabs for securing the sled.

bay tab.JPG

Then the coupler was glued into the airframe.

With the sled screwed into the coupler, a hole was drilled to serve as both a single vent port and also as access to the screw switch. A small drill bit was first used to puncture the outside of the airframe and make a mark on the sled where the screw in the switch should be placed. After removing the sled the airframe hole was enlarged to 1/8" for screwdriver access.

Next I glued on two brass plates to complete the screw switch around the marked location. One brass plate has a pre-soldered #4 nut. The other has an oversized hole for the screw so the circuit will not make contact until the circuit is screwed down. Both have been pre 'tinned' to make it easier to solder on wires.

switch 1.JPG switch 2.JPG

After gluing on the switch plates the g10 is drilled with a tap drill and tapped. I like this style of screw switch because it takes less room than commercial switches and it can be customized to where I want to put it. When soldering on wires don't overcook the plates or the epoxy may give up. Maybe this would be a good place for JB quick weld. In flight the two plates are screwed together and will maintain conductivity even if the epoxy were to crack.

End result is a screw switch exactly aligned with a hole in the body tube:

screw switch.JPG
 

Attachments

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
I like to secure batteries in a 'coffin' made this way:

IMG_6622Small.JPG

This seems like a great application for a vacuum bag.

final sled.JPG


Two igniter wires are routed from the Apogee and Main connections on the computer through holes on the bottom side of the sled, then up what would be the top edge of the sled in the picture above to be as far from the antenna as possible in this small package. After going through the holes in the top right they are routed through a hole in the end cap of the ebay and are plugged with plumber's putty. I use plumber's putty to seal the end cap onto the coupler.

As a little extra insurance I tied a lightweight Kevlar tether to one of the holes at the top left of the sled. This attaches to the friction fit motor. If the motor wanted to fall out it would help. If the screws securing the ebay wanted to pull out, it would help.

Here is a link to more electronics bays I have made with similar construction techniques. I have shortened the TeleMetrum antenna by adding a center loaded inductor. I have flown this configuration many times over the past few years. Recently I also characterized the impact on signal strength.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Nose Cone

The nose cone for eyelash was formed in a mold. Mold making and nose cone molding are shown below. Here are links to excellent videos showing the process of making molds and using molds.

mold making.JPG molding nose cone.JPG

For this rocket I used two layers of 3.7 oz S glass and a 1.43 oz outer layer laminated together with Aeropoxy. I didn't use a balloon or peel ply.

I like to put on the 1.43 oz layer and let it get stiff but not fully hardened before adding the thicker layers and sticking the two halves together. I think it helps keep prevent gaps/bubbles/defects in the outer surface of the nose cone.

On this rocket I let the nose cone harden with the tip down. Extra epoxy runs down and make a nice solid tip. I had a little more epoxy than I wanted and ended up drilling out some of the tip to make additional room for nose weight. I have made a couple nose cones that hardened tip up. This lets the extra epoxy drip out and leaves the tip hollow for nose weight, however I have had to fill in and form the tip afterwards.

This nose cone has an outer diameter larger than the OD of the rocket tubing. The plug was a standard VK nose cone sized for normal tubing. I glued a length of my body tube into the nose cone and tapered the diameter.

glue 1.JPG glue 2.JPG

This gave a slightly honest john shape to the rocket. It also provides a nose cone OD that matches the OD of the ring at the bottom of the body tube. This makes a good fit in the launch tower at both ends of the rocket. The coupler also fits really well into the nose cone because it was sized for this tubing.

tether 1.JPG tether 2.JPG

For the nose cone tether attachment, kevlar was frizzed out at the end and taped in to be glued along with the nose weight. The knot in the kevlar is a little extra insurance in case one side of the kevlar ever snaps. I size the tether attachment loop so that it is slightly inside of the bottom of the nose cone. This means that any rubbing and wear degrades the removable tether, not the permanent portion glued into the nose cone.

nose mass.JPG nose glue 1.JPG nose glue 2.JPG

208 grams of lead was glued into the nose cone. I hammered on fishing sinkers to make them conform to the profile of the nose and take up less space.

Launch mass of the the rocket without motor for the first two flights was 447 grams, so the 215 gram nose weight (lead + glue) represents about half the mass of the rocket.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Parachute

I made a 16" flat hexagonal parachute for Eyelash out of 0.5 Oz rip stop with sewn in reinforcement.

parachute.JPG

This post describes design and construction of some hexagonal parachutes. I use them when space and weight is a premium. I did not characterize the CD of this parachute with the car. I chose 16 inches for the diameter based on past flights with rockets of similar descent mass. I wanted a descent rate of about 35 ft/second because it is main at apogee, durable, and lightweight.

I used 100 lb kevlar for the parachute lines. A #5 ball bearing swivel rated for 230 lbs allows the parachute to spin.

recovery.JPG

I sewed a small deployment bag out of a nomex/kevlar blend. A sewn bag allows protection with less fabric/volume/mass. The bag has a diameter slightly larger than the rocket and is sewn only about half way up the side. I leave an overlapping tab that can snug up the parachute bundle when it is inserted. Once the bag deploys the top opens and the parachute has plenty of space to easily slide out. I sew a kevlar attachment loop onto the bag so it can hook to the tether right up by the nose cone.

750 lb kevlar was used for the rocket tether. There is no swivel at either end.

Here is a diagram of the layout of the entire recovery system:
recovery system.JPG recovery system picture.JPG
 
Last edited:

Neutron95

Well-Known Member
Joined
Jan 17, 2010
Messages
194
Reaction score
23
Do you have any more details on how you sized and rigged the deployment bag? I'm considering making my own parachutes for a 29mm high performance rocket, and space is at a premium.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Deployment Bag by request

The fabric for the deployment bag was cut out of a large nomex/kevlar chute protector I got from Wildman. The bag was made from two pieces of fabric. The main body is a rectangle with an extra protruding rectangle to be the tab. The bottom is a circle.

bag 1.JPG

To size the length of the fabric rectangle I Z folded the parachute and made the bag about an inch longer. To size the width of the fabric I measured the circumference of the body tube and added maybe 1/2". Some of this is seam allowance and some is a little extra diameter so when the bag is freed from the rocket it expands creating a loose fit around the parachute.

To size the fabric tab I wanted it to wrap about half way around the bag and to cover an unsewed opening that is about 1/3 of the bag length.

To size the circular bottom I cut a circle with a radius about 1/4" larger than the rocket diameter. Just set the tube on it and draw around.

First I sewed a zigzag stitch around the edges of the main body with nomex thread. I skipped the bottom edge. This stitching will help prevent fraying. If you have a serger and four spools of nomex that is even better on the edges. It is what is on professional protectors.

Next I folded the body in half, aligned the two edges, and sewed with nomex thread up the edge from the bottom to a smidgen beyond the tab. Then I positioned the circle on the bottom end and carefully sewed around the circular edge with nomex. I went around twice. This can be tricky.

bag bottom.JPG

Finally, for the attachment loop I took a length of kevlar, folded it in half, and sewed the two ends to the bag with kevlar thread.

To rig the bag I Z fold the parachute and slide it into the bag.

bag with Z chute.JPG bag rolled.JPG

Then I wrap the tab around the bundle snugly. If I take my finger off the bag pops open. The body tube keeps the tab/bundle from unwrapping until deployment. This rocket is 38 mm. I have made and used these little bags for 29 and 24 mm rockets also.

I attach the bag to the tether at the nose cone. I hooked it to a loop in the tether so it cannot slide. I attach the parachute down near the electronics bay. The attachment points are much further apart than the length of the parachute lines so that the rocket/nose cone tug on the parachute until it is out.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Motor prep: how not to make a tailcone

The simulations showed a modest altitude improvement from inclusion of a tailcone, so I made my first tailcones. The I216 does not have a separate closure. The reload includes the closure, so tailcones were bonded directly onto the reload closure.

I made a tube out of JB quick weld and fiberglass for the inside of the tailcone. I glued these around the nozzle with JB. The center tube had a 23 mm diameter and was 17 mm deep: an aspect ratio that was designed to avoid the Krushnic effect.
jb core.JPG

mandrel motor.jpg
I used hole saws to make rings out of styrofoam that fit around the center tube. The styrofoam was shaped with files and sandpaper. Cutting while pulling the file, rather than pushing, was more gentle on the styrofoam. I made troughs in the styrofoam matching the shape of the closure.

foam.JPG

Then I filled in some gaps with Rocketpoxy and glued pieces of 1.43 oz fiberglass over the tailcone with Aeropoxy.

glass 1.JPG finshed cone.JPG

When the first layer was hard I sanded it and added a second layer. This sanded and polished up well.

Please note that the tailcone addition was purely additive. Nothing in the original motor, or hardware was modified. It was simply a fiberglass and styrofoam shell glued to the stock motor.

I also want to stress using extreme caution when gluing a tailcone onto a reload or motor containing propellant.


The tailcones may look good and they are lightweight, but they don't work. I will explain in the flight section why this is a bad way to make a tailcone.

Finally, I dumped the powder out of the forward closure and potted the front end of the closure so that there would not be any blow by gasses that could dislodge the friction fit motor.
 
Last edited:

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Flight

Eyelash was flown three times on July 17 at the OROC Summer Skies launch. We had a gorgeous low wind day, perfect for main at apogee.

The first flight went to 14551'. It had a little kick into the wind coming out of the tower and landed, 401.5 seconds later about 900 feet from the pad.

GraphLotsOfSignals.JPG Statistics.JPG

I was a little disappointed because simulations had my expectations closer to 16k'.

three cones.JPG

Inspection of the tailcone, left one, showed that it had been mostly burnt away leaving only about 1/4". JB Weld didn't hold up. Maybe firebrick or metal next time. If 1/4" of tailcone is all that survives, I decided to trim off the rest. There is no point in sending it up if it burns off, consuming energy and adding friction at the same time.

tailcone reduction.jpg

The possibility of reducing the tailcone at the launch was a planned contingency. Calculations of the center of pressure with and without the tailcone were made prior to the launch. Because the tailcone will shift the Cp, both Cp marks were on the rocket to ensure safety in either configuration.

cp with tailcone.JPG
Flight #2 did better: 15220 feet with 431.6 seconds in the air. It came down about 2.4 miles away. Thank you GPS.

Graph1ManySignals.JPG Statistics.JPG

For flight #3 I made two changes:

I decided that there might be more drag on the rocket than was used in simulation, so I opened up RASAero II and degraded the paint finish until simulated results roughly matched the flight. Then I predicted the optimal mass and discovered that an extra 50 grams of mass would probably help counteract the extra friction. I tied about 56 grams of tungsten onto the ebay tab with kevlar.

On the tailcone I not only trimmed it, but I ripped out the styrofoam and the remaining short JB weld tube. These burned up on flight #2. All that remained was the ~1/4" long outer shell. It more or less survived flight #3. It's not clear that such a short tailcone adds benefit.

Flight #3 went to 15515 feet and stayed up 402.2 seconds, landing a little over a mile away. The maximum speed was between Mach 1.5 and 1.6. The peak acceleration was about 25 gs. The statistics below incorrectly list it at double that value due to a fleeting noise spike at takeoff.

Graph1BaroGPSSpeed.JPG Statistics.JPG Flight3D.JPG

Eyelash had a pretty good day for altitude and flight time: 45286' and 1235 seconds aloft, nearly 21 minutes.

I have an application pending for the Tripoli I class record.

I think that Eyelash may go higher if the fins are trimmed a little. The rocket was actually over stable at launch.

So that's how I made Eyelash: an altitude seeking minimum diameter I class scratch build.
 
Last edited:

SpaceManMat

Space Nut
Joined
Dec 20, 2013
Messages
660
Reaction score
47
Nice work kelly, congrats on the record. I’m curious about Tripoli’s take on the TC addition to the motor, specifically how much can be changed before it is considered a modification to the motor and hence no longer a commercial reload.
 

JLebow

Well-Known Member
Joined
Oct 31, 2017
Messages
75
Reaction score
120
Nice write up. It is great to see all the details freely shared. Congrats on the successful flights. Clearly you did your research, and the altimeter confirms it.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Paint and Surface Finish

Despite the natural look, Eyelash did have a careful polished paint job.

The paint was a gloss clear coat of VHT High Temperature Engine Enamel. It claims to be good to 550 F. This paint is a mixed bag. I love the final finish. It is hard, durable, and polishes up well. It is thin and lightweight compared to other finishes. You can also get this product in colors.

You need to be careful about applying the paint. Follow the recommendations on the can or it will run. "To avoid runs and sag, apply in 2 light coats, followed by one medium wet coat. Apply all coats within 1 hour, allowing 10 minutes between each coat." If you want additional coats you have to wait 7 days before applying them.

finish.JPG

Once the paint was fully dry, the rocket was lightly sanded to 1500 grit. Then automotive rubbing compound, polishing compound, and a coat of wax made it shiny and slippery.
 
Last edited:

TXWalker

Active Member
Joined
Dec 10, 2019
Messages
29
Reaction score
5
Location
Texas
I've been following your build thread. I really admire your workmanship. I'm planning a similar min diameter project in the future.

Could you elaborate more on your recovery solution. This is one on the areas I have not wrapped my mind around. It looks like you did a single deployment at apogee. Is this correct? I was thinking of doing a dual deploy with a long streamer at apogee and a parachute at 500ft. The streamer would be primarily for visual tracking. It adds about 4 inches to the length and wondered if it was required or not.
 

kbRocket

Well-Known Member
TRF Supporter
Joined
Nov 14, 2017
Messages
108
Reaction score
101
Location
Portland, OR
Could you elaborate more on your recovery solution.
I have received a few questions about the rocket, so yesterday I made clarifying edits to a several of the posts on this thread. The Parachute section now contains an overview diagram of the whole recovery system.

You are correct that the rocket used main at apogee. I configured the TeleMetrum for redundant apogee charges so that there was a primary and a couple seconds later a backup if needed. I used one #2-56 shear pin to secure the nose cone.

Most of my rockets are dual-deploy. Normally I wouldn't try for 15k' feet without it, but the goal of this rocket is altitude, not a general purpose flier. A longer rocket will take a little off the maximum altitude, however main at apogee may complicate recovery by allowing the rocket to drift a long way. You also don't want to drift outside the flight cylinder. You will need to assess what is safe and best for your site and situation. Learn to get "upper level" wind data for your launch site. The winds at these altitudes are typically much stronger and not always in the same direction as lower level winds.

Alternatives to dual-deployment are using a Jolly logic chute release or cable cutter. These seem promising, but also add more potential failure points. I experimented with putting a Jolly Logic chute release into this 38 mm airframe. I found a way to get it in there, but I decided to leave it out because it did not always release when used with such a small parachute. When the pin is pulled at a steep angle it can jam.

For the dual-deployment descent from apogee I have used, and watched other people use, small drogue chutes, streamers, or nothing (drogueless). I normally use a small chute. A chute or streamer is not required, but does give benefit beyond visual tracking. Some drag can help keep the front and aft sections of the rocket falling nicely. Without any drag the halves of the rocket can flat spin and rotate, twisting up the tether. Some rockets spin and some don't. I prefer adding enough drag such that the halves of the rocket are vertically oriented rather than flat.
 
Last edited:

gtm3323

Well-Known Member
Joined
Jan 20, 2009
Messages
248
Reaction score
9
Location
MN
Awesome flight and project Kelly! Thank you for sharing.
 

n27sb

N27SB
TRF Supporter
Joined
Jan 30, 2015
Messages
383
Reaction score
141
Nice work kelly, congrats on the record. I’m curious about Tripoli’s take on the TC addition to the motor, specifically how much can be changed before it is considered a modification to the motor and hence no longer a commercial reload.
Looks like a great flight. I too am surprised to find that Tripoli allows the addition of glued on parts to commercial motors.
 
Top