L3 Success: Compulsion, 4-in Fiberglass Frenzy XL build

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Could call Mike and ask him to ship you a new one....
+1 for that. Every HPR supplier I've worked with (heck, and Estes too) has had really good customer service, generally of the "oh, that problem? I'll fix it, no worries." variety.
 
I was marking lines for rail buttons last night, and in doing so, noticed that the fin slots do not appear to be well aligned with the booster tube. It appears that both the forward and aft fin slot were cut at the same time and thus both have the same misalignment. It's very hard to measure given the tools I have, an extruded aluminum level that seats on the outside diameter of my tube. However, I've been consistently measuring an alignment offset on all 3 (6) fin slots of roughly 0.025 to 0.030" from the bottom of the aft fin to the top of the forward fin (14.25 inches total fin length). If this is true, then I calculate that the rocket will rotate 1 revolution per every 501 feet of altitude. At it's expected top speed on an M1315W, ~ 1450 ft/sec, that would result in a spin rate of 2.9 revolutions per second. I'm not happy with that.

Rough calculation:
altitude per revolution ~= (tube circumference [radial in/rev]/fin offset [radial [radial in])*(fin length [in]) = (PI*4.03/0.03)*14.25 = 6014 in/rev * 1 ft / 12 in = 501 ft/rev

I plan to verify my measurement of this issue tonight, using whatever means I can find around the house. It's hard finding flat straight surfaces.. and I sure as heck don't trust my door frames.

I'm thinking that to correct this misalignment, I'll have to file the edge of my fin slots.. and thus have a very sloppy fit. Then I thought I could use a thin layer of rocketpoxy to build up the edge with the gap such that it would then produce a snug fit in the proper alignment. Sounds like a lot of work. Not happy with having to do this.. but I don't want my camera spinning at 180 rpm during the flight.
How are you able to measure the divergence to better than 1/40"? I've been using an 1/8" aluminum angle with 2" arms (I previously had a 1/16" with 1" arms that was not nearly good enough) and I'd be pretty dubious about it holding true to 1/40" over a 15" distance.

It seems the most accurate method would involve casting a laser (i.e. the ones used for drawing straight lines on walls) down the length of the tube and measuring divergence from it? You could force alignment of the laser and tube by requiring both to be level.

All that being said, I did notice that my frenzy's fin slots were not equally spaced around the tube (off by perhaps 1/8"). I believe I also noticed they might not be perfectly straight, but that the divergence was less than my ability to measure confidently let alone my ability to fix, and that they were almost certainly straighter than if I had tried to do the fin slots myself. They definitely pass the test of if I look down the rocket using the fin like a gunsight it looks like the fin intersects the tip of the nose.
 
I was marking lines for rail buttons last night, and in doing so, noticed that the fin slots do not appear to be well aligned with the booster tube. It appears that both the forward and aft fin slot were cut at the same time and thus both have the same misalignment. It's very hard to measure given the tools I have, an extruded aluminum level that seats on the outside diameter of my tube. However, I've been consistently measuring an alignment offset on all 3 (6) fin slots of roughly 0.025 to 0.030" from the bottom of the aft fin to the top of the forward fin (14.25 inches total fin length). If this is true, then I calculate that the rocket will rotate 1 revolution per every 501 feet of altitude. At it's expected top speed on an M1315W, ~ 1450 ft/sec, that would result in a spin rate of 2.9 revolutions per second. I'm not happy with that.

Rough calculation:
altitude per revolution ~= (tube circumference [radial in/rev]/fin offset [radial [radial in])*(fin length [in]) = (PI*4.03/0.03)*14.25 = 6014 in/rev * 1 ft / 12 in = 501 ft/rev

I plan to verify my measurement of this issue tonight, using whatever means I can find around the house. It's hard finding flat straight surfaces.. and I sure as heck don't trust my door frames.

I'm thinking that to correct this misalignment, I'll have to file the edge of my fin slots.. and thus have a very sloppy fit. Then I thought I could use a thin layer of rocketpoxy to build up the edge with the gap such that it would then produce a snug fit in the proper alignment. Sounds like a lot of work. Not happy with having to do this.. but I don't want my camera spinning at 180 rpm during the flight.

How are you able to measure the divergence to better than 1/40"? I've been using an 1/8" aluminum angle with 2" arms (I previously had a 1/16" with 1" arms that was not nearly good enough) and I'd be pretty dubious about it holding true to 1/40" over a 15" distance.

It seems the most accurate method would involve casting a laser (i.e. the ones used for drawing straight lines on walls) down the length of the tube and measuring divergence from it? You could force alignment of the laser and tube by requiring both to be level.

All that being said, I did notice that my frenzy's fin slots were not equally spaced around the tube (off by perhaps 1/8"). I believe I also noticed they might not be perfectly straight, but that the divergence was less than my ability to measure confidently let alone my ability to fix, and that they were almost certainly straighter than if I had tried to do the fin slots myself. They definitely pass the test of if I look down the rocket using the fin like a gunsight it looks like the fin intersects the tip of the nose.


I have a heavy duty 6 foot long aluminum level, that has a deep "I" cross section. It is straight, and relatively stiff/resistant to torsional deflection. I found that if I remove the center bubble level and lay the level sideways on my tube, it will sit nice an flat, and lay parallel with the axis of the tube nicely. I used this to mark a line between fin slots from end to end on the booster tube. I'm confident in this method since I'm able to turn the level over or flip in around and it will always line up perfectly on this line again. If I intentionally put some offsetting force on one side of the level vs. the other while holding it down in the center on the tube, it will deflect perhaps 1/16" over the full length. It fits good enough that it's unlikely that I'm misaligning the level.

I then lined up the level from the back of the aft fin, and saw that the front of the forward fin was offset by a small but measureable amount.. by my eye I'd guessed 1/32". So, I set my calipers to 0.030 inches, and made a little mark on the forward side of the forward fin slot and verified it with the level.. it was very close.. perhaps a little less than 0.030". That's where I got my number.

I do have a "sheet" laser that I use around the house.. it has auto leveling horizontal and vertical sheets or lines that it can project. Your suggesting is interesting.. I'll have to think of how I can use that to confirm my findings.
 
I have a heavy duty 6 foot long aluminum level, that has a deep "I" cross section. It is straight, and relatively stiff/resistant to torsional deflection. I found that if I remove the center bubble level and lay the level sideways on my tube, it will sit nice an flat, and lay parallel with the axis of the tube nicely. I used this to mark a line between fin slots from end to end on the booster tube. I'm confident in this method since I'm able to turn the level over or flip in around and it will always line up perfectly on this line again. If I intentionally put some offsetting force on one side of the level vs. the other while holding it down in the center on the tube, it will deflect perhaps 1/16" over the full length. I fits good enough that it's unlikely that I'm misaligning the level.

I then lined up the level from the back of the aft fin, and saw that the front of the forward fin was offset by a small but measureable amount.. by my eye I'd guessed 1/32". So, I set my calipers to 0.030 inches, and made a little mark on the forward side of the forward fin slot and verified it with the level.. it was very close.. perhaps a little less than 0.030". That's where I got my number.

I do have a "sheet" laser that I use around the house.. it has auto leveling horizontal and vertical sheets or lines that it can project. Your suggesting is interesting.. I'll have to think of how I can use that to confirm my findings.
Sounds reasonable. I'd make triple sure there isn't small amounts of crud or residue though on the tube or level before going to extremes to fix the problem. My body tube certainly had little bits of crud on it and any of that along the 6 feet that the level contacts the tube could easily shift things by the amounts we are talking about. I mention it because I noticed this stuff on my tube after being annoyed that my lines down the tube didn't pass my "gunsight" test :) I guess what makes me skeptical that I personally could achieve this accuracy is my observation that if I slid my angle around the tube by applying force only on one end of it I was able to get it to rest in slightly different angles. Not much, but definitely more than 1/40" over 15"!
 
Sounds reasonable. I'd make triple sure there isn't small amounts of crud or residue though on the tube or level before going to extremes to fix the problem. My body tube certainly had little bits of crud on it and any of that along the 6 feet that the level contacts the tube could easily shift things by the amounts we are talking about. I mention it because I noticed this stuff on my tube after being annoyed that my lines down the tube didn't pass my "gunsight" test :) I guess what makes me skeptical that I personally could achieve this accuracy is my observation that if I slid my angle around the tube by applying force only on one end of it I was able to get it to rest in slightly different angles. Not much, but definitely more than 1/40" over 15"!

My tube is 100% free of "crud" but does have plenty of sharpie writing on it. I use liberal amounts of 91% or 99.9% IPA with fresh paper towels to wipe the tube off whenever I've been working with epoxy.

I originally started with an aluminum angle to measure down the tube.. but I found that because it was not very stiff in torsion, that it would not produce reliable/repeatable results. So, I tossed that idea and searched for something better.. that's when I found my heavy duty 6 foot level could be used with significantly better results.
 
Sanded bonding surfaces using fresh 80 grit sand paper:
Motor Tube Centering Rings:
Roughing Up Motor Mount Surfaces for Bonding IMG_3404.JPG Roughing Up Motor Mount Surfaces for Bonding IMG_3406.JPG

Exterior of Motor Tube where Fins will be bonded:
Roughing Up Motor Mount Surfaces for Bonding IMG_3407.JPG

Interior of Booster Tube along side of all fin slots.. This took some time... I had to go over the sanding 2 or 3 times until I was satisfied in the scoring:
Roughing Up Booster Tube Surfaces for Motor Mount and Internal Fin Fillet Bonding IMG_3394.jpg
I used 99.9" Isopropyl Alcohol to clean the surfaces before bonding:
Clean Motor Mount Tube and Centering Rings with IPA IMG_3419.JPG

Motor Tube is in, and the Flourescent Red Pigment RocketPoxy is very visible through the Fiberglass:
Motor Mount Tube Bondedinto Booster IMG_3420.JPG

Here's a shot of the nice fillet on the forward Centering Ring with the Kevlar Anchor Straps on either side:
Forward Centering Ring Nice Fillets IMG_3428.JPG
 
I installed two 8-32 PEM nuts for use with Rail Button Mounting. I plan to apply a fillet of RocketPoxy around the upper PEM nut to prevent it from snagging any recovery gear. They protrude approximately 0.100 " internally, and are flush to the exterior.

I had an issue with a PEM nut creating some stress fractures in the G12 tube near the base.. I think it's related to the fiberglass filament angle being too shallow near the end of the tube. You can see the stress fractures in the glass.. and see how they are in the direction of the fiberglass filament wind. The filament wind is normally something like +45/-45 degrees in the tube, but near the bottom it is significantly reduced, and thus limits the tube's ability to carry load in the axial direction. I'm guessing that the tube fabrication winding process uses a machine that spins fiberglass filament over the tube being created by making passes back and forth over a spinning tube, and that the bottom of this booster tube was the section of the tube that was near the end of travel for the tube fabrication process. I just checked the booster tube for my 4" DX3 XL and it has the same reduced wind angle near the bottom.

Aft PEM nut installation near bottom of booster tube:
Aft PEM Nut External View of Stress Cracks through Reduced Angle Fillament Wind IMG_3417.jpgAft PEM Nut Internal View of Stress Cracks through Reduced Angle Fillament Wind IMG_3414.jpg


The PEM nut I installed in the mid section of the booster went in just fine without any stress cracks in the G12 Filament Wound Fiberglass. The Filament wind angle is roughly +45/-45 in that location as it should be.
Forward PEM nut installation:
​Forward PEM Nut for Rail Button IMG_3580.JPG
 
I measured the fin slot misalignment to be approximately 0.028 to 0.034" across the length of both fin slots. Here's a photo that shows how I measured this.. you may need to open the photo, zoom in and pan from side to side to see how the fin slot "runs out" about 0.034".
Fin Slots Misalignment IMG_3185.jpg Fin Slots Misalignment IMG_3191.jpg

I decided that I could straighten the fin slots by carefully widening both sides of the fin slots. I used 80 grit paper glued to a tongue depressor and sanded the fin slots. For each pair of find slots, I blended the amount of material removed across the length of both fin slots The sanding removed approximately 0.015" on one end and carefully tapered to no material removed on the other end of the fin slot pair. This technique minimized the final corrected fin slot width.
Fin Slots Misalignment Correction IMG_3182.jpg

I then created a fin alignment jig using the flattest plywood that I had in my garage.. which was very flat. I use my table saw to cut two nice square lengths of 3/4" plywood, and cut a 30 degree bevel on one of the lengthwise edges of each piece. With the beveled sides down and opposing each other, the two pleces of plywood will straddle the diameter of the tube and align straight down the tube. Using a flat surface and a T-square, I was able to tighten the C-clamps over the plywood with 2 aft fins creating a gap just wide enough for the forward fins to slide into. I was careful to get the alignment just right before clamping the jig together. As far as I can tell, it has worked extremely well, and it's alignment down the tube is in perfect agreement with the extruded aluminum I-section from my long bubble level.
Fin Alignment Jig IMG_3460.jpg Tack Fins In Alignment IMG_3484.jpg
 
I was not happy with the stock bevel that was on each of the fins. I made a simple sanding jig and used 60 grit, then 180 grit sand paper to improve the fin bevels on the forward fin leading edges and the aft fin trailing edges. Here's some photos:

Fin Beveling Setup:
Fin Bevels Improvement IMG_3443.JPG

Forward fin bevel improvement, Improved Bevel is on the Left, Stock Bevel is on the Right:
Fin Bevels Improvement IMG_3440.JPG

Aft fin bevel improvement, Improved Bevel is on the Left, Stock Bevel is on the Right:
Fin Bevels Improvement IMG_3436.JPG

All Bevels Done.. they turned out pretty good:
Fin Bevels Improvement IMG_3447.JPG
 
Time to tack down the fins... first I sanded each of the fins near the root where it will bond to the motor and booster tube using 80 grit paper.

Sanded Fins:
Fin Bevels Improvement IMG_3458.JPG

Just a small amount of RocketPoxy with Flourescent Red Pigment to Tack down each fin:
RocketPoxyMix with Flourescent Red Pigment IMG_3474.JPG Tack Fins In Thin Bead of RocketPoxy IMG_3476.JPG


I used the fin alignment jig to be certain that the fins were aligned:
Fin Alignment Jig IMG_3460.jpg Tack Fins In Alignment IMG_3484.jpg

A small fillet formed where the Rocketpoxy pressed out from between the fin and motor tube on each fin. The Rocketpoxy on the sanded and cleaned fiberglass had excellent wetting properties. The Rocketpoxy wicked outward and up from the bond area. A much larger fillet will be added to this joint after all fins have been "tacked" down.
Assuming Rocketpoxy has a fiberglass-to-fiberglass lap shear strength of ~2000 PSI, these "tacked" on fins have a very strong bond to the motor tubes even without the final reinforcing fillets. I think the largest load these fin bonds will need to endure will be the shock from landing on a hard surface... such as when the full landing shock is imparted on a single aft fin.
Tack Fins In Minimal Fillet IMG_3515.JPG Tack Fins In All In Seen Through Motor Tube IMG_3514.JPG

This took a couple of days to complete, since I mixed one batch of Rocketpoxy per fin, and tacked down one fin at a time, then allowed 8 hours for the epoxy to cure sufficiently before working on the next fin.
 
I injected the internal fillets through the aft end of the rocket using a syringe and a length of 3/8" polyethylene tubing. After a short "learning curve" I was able to inject these evenly and I am very happy with the result. Here are a few pictures of the process:

Drawing Rocketpoxy into tube using vacuum from syringe. The syringe is reusable as long as it doesn't get epoxy inside of it during the process. I completed all 12 fillets using one syringe.
IMG_3546 Internal Fillets.JPG


Injecting Rocketpoxy fillets through the aft of the rocket. Notice the nice thick line of epoxy that has been injected along the first half of the forward fin:
IMG_3543 Internal Fillets.JPG

Close-up of injected epoxy inside booster and the shadow of the injection tube as I injected a healthy line of Rocketpoxy. I used a flashlight inside the rocket to make the tube more visible during this process:
IMG_3544 Internal Fillets.JPG


After the epoxy was injected, I then tilted (rolled) the rocket from side to side to distribute the epoxy, forming fillets. I was careful to level the fin after this step, so that the epoxy would flow back to level before curing.
IMG_3565 Internal Fillets.JPG


The first internal fillet was a little sloppy, but I quickly got the hang of this and was able to lay down some nice even fillets. The trick was not to inject too much epoxy near the edge of the fin, and keep on top of the flowing epoxy until the flow rate subsided as the epoxy cured.
Here you can see how well these internal fillets turned out:
Nice uniform thick fillets can be seen through the side of the booster:
IMG_3561 Internal Fillets.JPG

Nice thick fillets are visible through the inside of the motor tube:
IMG_3562 Internal Fillets.JPG

I backed up the camera to get shot that captures the thickness of all 3 internal fillets:
IMG_3588 Internal Fillets Thickness Aft view.JPG
 
I installed two 8-32 PEM nuts for use with Rail Button Mounting. I plan to apply a fillet of RocketPoxy around the upper PEM nut to prevent it from snagging any recovery gear. They protrude approximately 0.100 " internally, and are flush to the exterior.

I had an issue with a PEM nut creating some stress fractures in the G12 tube near the base.. I think it's related to the fiberglass filament angle being too shallow near the end of the tube. You can see the stress fractures in the glass.. and see how they are in the direction of the fiberglass filament wind. The filament wind is normally something like +45/-45 degrees in the tube, but near the bottom it is significantly reduced, and thus limits the tube's ability to carry load in the axial direction. I'm guessing that the tube fabrication winding process uses a machine that spins fiberglass filament over the tube being created by making passes back and forth over a spinning tube, and that the bottom of this booster tube was the section of the tube that was near the end of travel for the tube fabrication process. I just checked the booster tube for my 4" DX3 XL and it has the same reduced wind angle near the bottom.

Aft PEM nut installation near bottom of booster tube:
View attachment 327246View attachment 327247


The PEM nut I installed in the mid section of the booster went in just fine without any stress cracks in the G12 Filament Wound Fiberglass. The Filament wind angle is roughly +45/-45 in that location as it should be.
Forward PEM nut installation:
​View attachment 327265


Following up on this issue...
Even with the cracks in the G12, the aft PEM nut installation was very solid. I debated just leaving it as-is, or doing something to reinforce the tubing. I decided to lay down a few layers of 0-90 fiberglass inside the booster around the stress cracks, making sure to get plenty of fibers oriented longitudinally across the stress cracks. I used Aeropoxy ES6209 as the matrix epoxy. While not exactly pretty.. the result was excellent. The PEM nut is captured in the epoxy glass and the internal threads are clean of epoxy.
Here's a picture:
Aft PEM Nut Crack Reinforcement IMG_3639.JPG
 
I used a 1 1/4" PVC pipe coupler with an OD of approximately 1.5 " to pull and form the Rocketpoxy fillets for all 12 fins. This took a number of days to complete, given the number of fillets and the time required for the Rocketpoxy to stop flowing. They turned out excellent and are nice and uniform from fin to fin. I opted to fillet around the ends of the fins, hoping to add a little more strength to counter landing impact stresses on the fins.
Here are some pictures:

Sanded with 80 grit.. again to be certain of a good bond:
IMG_3591 Fin Fillets.JPG

Clean with 99.9% IPA just before applying Rocketpoxy:
IMG_3592 Fin Fillets.JPG

Initial "pour":
IMG_3595 Fin Fillets.JPG

A bit messy:
IMG_3598 Fin Fillets.JPG

Keep it level after pulling radius numerous times using the PVC coupler:
IMG_3600 Fin Fillets.JPG

Resulting fillets were very nice:
IMG_3604 Fin Fillets.JPGIMG_3602 Fin Fillets.JPGIMG_3646 Fin Fillets.JPG
 
I installed the Aft centering ring, with 3 of 4 fillets. I did not form a fillet where the Aeropack 75mm motor retainer will butt up against the centering ring:
IMG_3640 Aft Centering Ring.JPG

After sanding and cleaning the motor tube, I put a thin 100% coat of JB weld around it's circumference:
IMG_3648 Motor Tube with JB Weld.JPG

Per the Aeropack instructions, I used Acetone to clean the motor retainer before applying a thin uniform coat of JB Weld around its circumference:
IMG_3650 Retainer with JB Weld.JPG

I failed to take a photo of the installed motor retainer.. but you get the idea..
 
For the electronics in the nose cone, does the wood brace against the side of the cone at the top or is it only touching at the base? If it only touches at the base, that is quite the long lever arm so I'd try some drop tests to make sure it'll survive hitting the ground at various angles under chute. The nose mounted tracker housing ive been working on wasn't strong enough at first and broke onlanding. And it was much much shorter than yours, I was pretty surprised it broke. I've built it now to survive free fall from 7 feet or so. You may be fine though, mine is built to be lightweight.

Also curious why the GPS mounting board is so long! I have also experienced failure so I ended up with a short piece of G10.
View attachment 319592

You two saw that issue right away! Yes, that plywood is cantilevered from the base that it's embedded in. It will deflect significantly in bending across the flat side if I impact the nosecone in that direction. The plywood sled will not break even if the end impacts the side of the nosecone, but I don't think the tracker will survive these impacts. I have not bonded the coupler in the nosecone yet, as I want to be confident that I have a good solution to this issue first. I have in the past simply placed foam just below the tracker between the nosecone and the plywood sled to resist displacement. This exact tracker and long mounting sled have flown twice in my other 4" FG rocket which has the same mounting ring in it's nosecone. The foam solves the bending problem.. but it's not the clean solution I hope to achieve.

The reason I made the plywood sled so long was to get the GPS antenna out from underneath 2 bonded layers for filament wound fiberglass. I figure that there may be better reception under just a single layer. Also, the tracker can easily be repositioned on the sled. So, when I cut the sled I cut it long knowing that I could make final adjustments later.

Following up on the long altimeter sled bending concern..
I decided to add a support inside the nosecone. I did some testing and where I have the support mounted prevents the lower bending modes of the plywood sled. Drop tests on the side produce no low frequency oscillations. However, I can feel the stock wire Eggfinder Antenna's first bending mode vibration.
Here are some pictures:

IMG_3617 Tracker Support Ring.JPG IMG_3618 Tracker Support Ring.JPG IMG_3625 Tracker Support Ring.JPG IMG_3626 Tracker Support Ring.JPG IMG_3736.jpg IMG_3734.jpg
 
I used 1/4" hobby plywood from Hobby Lobby and my table saw to quickly fabricate the first idea for an altimeter sled that popped into my mind.. it turned out nicely.
Here's some pictures:

Sketch of idea:
IMG_3663 Sled.JPG

Cut out parts, and drilled holes for the threaded rod using a 1/4" auger bit to prevent damage to the thin walls:
IMG_3662 Sled.JPG

Assembly:
IMG_3664 Sled.JPG IMG_3665 Sled.JPG

I also cut some screw switch mounting blocks at 30 degrees using the table saw. I mounted them using 3mm nylon screws after drilling and tapping the plywood mounting blocks.
Here's a picture with the screw switches installed:
IMG_3668 Sled.JPG
 
I soldered up an Eggtimer Quantum WiFi Altimeter. I test fired igniters and verified the barometer. It's working nicely. I'm not sure if I'll make this the main or backup altimeter.

IMG_3691 Eggtimer Quantum.jpg IMG_3693 Eggtimer Quantum.jpg IMG_3692 Eggtimer Quantum.jpg
 
Masking tape will make those fillets a lot less messier!

View attachment 328553

I thought I'd go without the tape this go-around. You're right, however, I would have saved a good amount of time not having to scrape and wipe the excess off the fins and body tube near the fillet. I'll use tape again next time...

is your fin can picture of a Nike smoke with molded fins? I like the green/yellow epoxy. Is that one of the Rocketpoxy pigments?
 
I thought I'd go without the tape this go-around. You're right, however, I would have saved a good amount of time not having to scrape and wipe the excess off the fins and body tube near the fillet. I'll use tape again next time...

is your fin can picture of a Nike smoke with molded fins? I like the green/yellow epoxy. Is that one of the Rocketpoxy pigments?

I don't think I've ever applied fillets without tape , but that's me. Anyhow rocket looks great!
That's a Madcow AGM Pike and yes, Rocketpoxy green pigment.
 
I will be using a Perfectflite StratoLoggerCF for the primary altimeter, and an Eggtimer Quantum for the backup altimeter.
I'll be using a 1S 750mAh LiPo for the StratoLoggerCF, and a 2S 460mAh LiPo for the Eggtimer Quantum. Missile Works screw switches will break the battery connection to both of these altimeters and thus the deployment charges.
I mounted both altimeters and their corresponding LiPo batteries in the upper half of the altimeter bay. The StratoLoggerCF and its battery are on one side of the altimeter bay sled, and the Eggtimer Quantum and its battery are on the other side.
I drilled and tapped the 1/4" plywood sled for use with 3mm nylon screws to mount the altimeters and screw switches.
For both altimeters, the wiring between the battery and the altimeter is 20 AWG Silicone, and the deployment output wiring is 22 AWG Silicone.
I crimped in some locking connectors inline for all deployment channels as well as the Amplified Beeper. This allows easy disassembly. I trust these connectors as they have worked nicely in my two most recent builds.
All wiring connections to the altimeter have been strain relieved using zip ties and heat shrink insulation as seen in the photos below:

StratoLoggerCF Integration Details:
IMG_3725.JPG IMG_3726.JPG

The additional 3 conductor wire is for a prototype amplified beeper. That white tube is not a charge cup.. it's a short section of tubing that helps to protect the beeper. The amplified beeper has good volume, and should be very helpful in locating the rocket in situations where line of sight to the rocket is obstructed.
IMG_3727.JPG

Eggtimer Quantum Integration Details:
IMG_3729.JPG IMG_3731.JPG IMG_3732.JPG

Side views:
IMG_3728.JPG IMG_3733.JPG


As I've only used the top half of the altimeter sled I'll have plenty of room to mount other devices such as a Raspberry Pi Zero etc. I have a miniature camera setup with USB audio for a Pi Zero, and am thinking about integrating accelerometers and perhaps a redundant pressure sensor. That will be a good winter project.
 
Time to configure and test the altimeters...

I configured the StratoLoggerCF to fire the drogue charge at apogee and the main at 1300 ft. I flashed the Eggtimer Quantum with software version 1.06n, then configured it to fire the backup drogue charge at 1 second after nose-over, and the main charge at 1100 ft.

I assembled the altimeter bay and verified that it has very little pressure leakage through the bulkheads. I then installed 4 Ematches and performed an altimeter ground test using a shop vac. I taped off the booster shear pin holes, payload fastener holes and 2 of the 3 vent holes. Then I used a shop vac to quickly draw down the pressure inside the altimeter bay through the open vent hole, waited for the drogue Ematches to ignite, slowly released the vacuum, and verified the main ematches to ignite.

The Ematches all ignited in the expected order. I downloaded the data from each altimeter and it looks as-expected. Test successful. Below are the data plots for this test from both Altimeters:

StratoLoggerCF Data:
20171002 StratoLoggerCF Vaccum Test Frenzy XL.jpg

Eggtimer Quantum Data:
20171002 Eggtimer Quantum Vaccum Test No1.jpg

Next up will be deployment charge ground testing...
 
Have you considered using an LED+resistor to simulate ematch ignition instead of burning live igniters?
Just something from broke college team days when our igniters were few and precious lol.

Love the build so far!
 
After careful consideration of venting, deceleration and shock loads, I have decided to go with three 2-56 nylon shear pins for the booster-to-altimeter bay joint, and three 4-40 nylong shear pins for the payload-to-nosecone joint. (Of course the shear "pins" are actually nylon screws.)

The quantity of 4F black powder required to shear the pins and separate the rocket sections was calculated using an online calculator. The calculator can be found here: https://hararocketry.org/hara/resources/how-to-size-ejection-charge/
The calculator was based on the Ideal Gas Law, PV = nRT.
The equation to find the quantity of 4F black powder in grams is:

Initial estimates for black powder charges were calculated using conservative values for the shear strength of both the 2-56 and 4-40 shear pins. As the shear pins are threaded, the actual cross sectional area of the screw that fails under shear stress can vary. To be conservative, the pitch diameter of the screws was used to determine the "maximum" shear strength per screw. For 6/6 Nylon, 10,500 psi was used for the shear strength calculations. Using this, the maximum shear strength is 46lbs for 2-56 nylon screws and 76 lbs for 4-40 nylon screws. An online reference for shear pin strength is https://www.feretich.com/rocketry/Resources/shearPins.html.
Given that there are three 2-56 pins in the drogue section, the force required to separate the drogue section is 138 lbs. Similarly, with three 4-40 pins in the main section, the force to separate the main sections is 228 lbs. The internal cross sectional area of the rocket is pi*r^2 = pi*(3.9 in/2)^2 = 11.95 sq in. Thus, the pressure required to separate the drogue section is P = F/A = 138 lbs/11.95 sq in = 11.5 PSI. Similarly, the pressure required to separate the main section is 19.1 PSI. The length of the drogue compartment is 19 inches, and the length of the main compartment is 13 inches. Using these values plus a 25% safety factor in the online calculator or above Ideal Gas Equation, the following quantities of 4F black powder will serve as starting points:
⦁ Drogue 4F Black Powder Charge: 1.7 grams
⦁ Main 4F Black Powder Charge: 1.9 grams
Based on experience with a similarly sized rocket, these charges look like a good starting point. The final charge quantities will be determined by ground testing.
 
Have you considered using an LED+resistor to simulate ematch ignition instead of burning live igniters?
Just something from broke college team days when our igniters were few and precious lol.

Love the build so far!

Funny you mention that.. I was thinking to myself that I should find some old Christmas tree lights or something.. since I burn 4 igniters per ground test.
I'll have to look into this... I have some pre-fabbed bright LEDs that I use with my FPV multirotors... perhaps they would work right out of the bag. I might have an issue with the deployment voltage from the StratologgerCF, since it's only using a 1S (~4.0V) LiPo. Hmmm. Thanks for the feedback.
 
Funny you mention that.. I was thinking to myself that I should find some old Christmas tree lights or something.. since I burn 4 igniters per ground test.
I'll have to look into this... I have some pre-fabbed bright LEDs that I use with my FPV multirotors... perhaps they would work right out of the bag. I might have an issue with the deployment voltage from the StratologgerCF, since it's only using a 1S (~4.0V) LiPo. Hmmm. Thanks for the feedback.

The high amperage from the firing channel will likely bake a standard LED (as in nothing but the plastic encased diode and terminals) and then you'll have created another consumable lol. We used something like a 100 or 200 Ohm resistor in series with the LED.

I never considered Christmas lights, but I have seen them used in a flier's test setup recently.
 
I've been debating how I'm going to do the vents for the parachute compartments. The Madcow instructions say to drill a single 1/8" diameter hole in the middle of the payload section, and a 1/8" hole in the booster section just forward of the top centering ring. I was thinking that 1/8" might be on the small side for this project, and thought perhaps a 9/64" hole might be better, particularly for the booster/drogue compartment given it's larger volume and smaller shear pins. While a single 9/64" diameter hole should work for both the payload and booster sections, I'm considering using 3 smaller 3/32" diameter vent holes for each section, aligned with the fins. It occurred to me that having 3 smaller holes would make it less likely that the rocket vent is blocked by recovery gear. Three vents would also give flow symmetry during ascent to apogee, rather than having a single larger "flow disturbance" on just one side of the airframe. If I do go with single vent holes, I would align them with the fin which is 180 degrees away from the rail buttons.
I plan to drill the payload/main vent(s) just below the nosecone bulkhead, and the booster/drogue vent(s) just above the forward centering ring.
Any opinions out there how best to vent these compartments? Thoughts on 3 smaller vents vs. 1 single larger vent?
 
I've used a single 3/32" hole on both compartments on my 4" Punisher.

Thanks Chris. I've been taking a close look compartment venting. It's interesting to consider how various scenarios drive the vent hole size requirements. Compartment Volume, Vent Hole Size and rate of climb (rate of pressure change) are the major factors.
 
I found a nice write up by David Shultz titled "Parachute Bay Pressure Vents" that describes how to model compartment venting using a simple system that behaves like and RC filter.
Apparently, Bob Krech had posted a formula for computing the time constant for an altimeter bay vent on Rocketry Online but it is lost due to a server crash.
The Altimeter Bay Time Constant formula is:
T = L*((D/d)^2))/4000
L is the length of the compartment
D is the inside diameter of the compartment
d is the diameter of the circular vent hole

The write up, see link above, goes on to show how a first order digital filter can be used to model this and provides a link to a nice little C program. While the original C program was written to work with RockSim, I modified it to work with exported data from OpenRocket consisting of time, altitude, velocity, pressure, and time step. The output is a csv text file. Here's a short sample of the output file:

#Compartment Diameter = 3.900000
#Compartment Length = 27.000000
#Vent Diameter = 0.132580
#TC = 5.840863
time (s), altitude (ft), velocity (ft/s), pressure (psi), filtered pressure (psi), pressure difference (psi), force (lbs)
0.010000, 0.000000, 0.000000, 14.700000, 14.700000, 0.000000, 0.000000
...
0.250000, 7.503700, 77.613998, 14.697099, 14.699971, 0.002872, 0.034314

I manually crop the output file after apogee deployment events at peak altitude.


The payload section of my Frenzy will need to vent the payload tube volume plus the nosecone volume. For vent calculations, I used L = 27 inches for the payload compartment, and D = 3.90 inches. After careful consideration, I've decided to use three 3/32 diameter vent holes in my Payload Compartment, with the holes just below the nosecone coupler, spaced 120 degrees apart and aligned with the fins to minimize flow disturbance over the altimeter bay vent holes. As I'm using 3 holes, I calculated an "effective diameter" for the three 3/32" vent holes using the following equation:
d = EffectiveVentDiameter = 2*sqrt((n*A)/pi)
n is the number of holes
A is the area for a single hole

This "effective diameter" is what I used to input vent hole diameter into the simulation program for multiple vent hole configurations.

Thus d = 0.16238" for n=3 3/32" vent holes

I ran the simulation for multiple scenarios, assuming that one or two of the three vent holes could be blocked. The vent diameters used where: d=0.16238 (3 holes), d=0.13258 (2 holes), d = 0.09375 (1 hole).
It's interesting to see how much additional force due to trapped internal pressure can be created by blocked vent holes. Here are my results for the Payload/Main Parachute Compartment where I've plotted the output/results using Google Sheets:
Pyld 332 holes F&AvT.png Pyld 332 holes FvA.png

I'll be using three 4-40 shear pins to secure the nosecone to the payload tube. They were sized primarily for potential shock load upon the kevlar cord snapping tight during the apogee drogue deployment, but they must also withstand the added stress due to any trapped pressure within the payload compartment. Three 3/32" diameter vent holes will be a good size and number for the Payload Compartment, even if one or two holes are completely blocked by recovery gear.

For the Booster/Drogue Parachute Compartment, I used L = 22 inches, D = 3.9 inches and d=0.16238 (3 holes), d=0.13258 (2 holes), d = 0.09375 (1 hole), where d is the effective diameter for the number of 3/32" diameter vent holes being simulated.

Bstr 332 holes F&AvT.png Bstr 332 holes FvA.png

I'll be using three 2-56 shear pins to secure the booster section to the lower avionics bay. These pins will combat separation forces due to internal pressure as well as drag deceleration forces after motor burnout. The three 3/32" diameter vent holes look to be a good size for the Booster Compartment. Similar to the Payload shear pins, the Booster shear pins should be able to tolerate one or two vent holes becoming blocked by recovery gear.I've plotted the output/results using Google Sheets.
 
Back
Top