Madcow FG 4" Frenzy XL build thread

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rocketsam2016

Well-Known Member
Joined
Aug 31, 2016
Messages
270
Reaction score
5
After eyeing this kit for some time I couldn't resist the good sale in November, and I'm ready to start building! I'd love to someday fly it on an M, but that won't be anytime soon since I want a lot more flights under my belt in the L2 range and MMMSC is limited to 10k feet.

I'm not making massive changes to the kit, but thus far I'm planning the following addendums/changes:

  • Use RocketPoxy everywhere (except motor retainer) and inject internal fillets. The injection process sounds like a mess to me, but so does doing them with a dowel between a 75mm mount and a 4" air frame with split fins and people seem to like the injection method so time to try it out.
  • Add two extra centering rings (making a total of 5) to make the dams for injection.
  • Aeropack 75mm retainer epoxied on with jb weld
  • 1" kevlar strap epoxied to two sides of the motor mount. The kit calls for attaching shock cord directly to motor, but I'll mostly be flying 54mm CTI which don't natively allow for that, and if I'm not adding a bulkhead might as well leave the motor charge in as a tertiary backup of the apogee split. I've been going back and forth for some time about the kevlar strap method vs adding a bulkhead, but (as with injecting) the kevlar strap method seems very popular so time to give it a shot
  • 2 eggtimer quantums in the avbay. TBD whether I add a physical switch, though I suppose it'd be required if I use this for L3
  • Eggfinder gps in the nose. I've already designed+built a twist-in enclosure to use the eggfinder in another rocket, so all I need is to cut a house-shaped hole in the nose cone bulkhead. More on this later.
  • Using the included ~30' TN shock cord for the booster -> avbay connection, and bought a 25' tubular kevlar cord to go from avbay -> nose. I'm on board with all the arguments against too-long cords, but I fly near trees and I want a shot at getting this back from one.
  • (tentatively) 60" fruity iris main and 15" fruity classic elliptical drogue. I've bought these but left them in the package in case the final weight is off from my estimates.
  • Using a fruity deployment bag with the main, probably with a 24" fruity elliptical for the pilot. I've used deployment bags before and I'm hooked.
  • I'm thinking about some sort of external mount for a couple of cameras. I don't mind the drag - this isn't a lean and mean rocket and I fly on the east coast, and I've got this crazy idea about doing either stereoscopic mounts for 3d or some sort of panorama.

Madcow lists 14lbs which I took as the dry kit weight, so I was assuming 17-18lbs or so fully built and loaded other than the motor. However, all the fiberglass parts and hardware supplied with the kit plus the aeropack and the parachutes only comes to ~13.25lbs, so even after I build, paint, build the avbay sled + electronics and add the gps, this may come out lighter than I expected. We'll see!
 
Excellent choice of rockets. The fin can pictured on the left is my 3" Frenzy XL using the same 5 CR technique you are planning to use ( the fifth CR is actually a thrust plate). On the right is the fincan for my Mini-FrenzyXL DD. Instead of using extra CR's and injected fillets like I did on the 3" I am using fin pockets. The fin pockets are the same technique I used on my L3 build and other rockets since. Basically the dams on each side of the fin are spaced to allow the size of fillet (internal fillet,that I want), then since they fit tight to the mmt and airframe both, the mmt is epoxied in to the airframe with the space between the dams lined up with the slots in the airframe. I can then use thicker epoxy that is inserted between the dams, then the fins are inserted and the excess epoxy that squeezes out is wiped off the exterior of the airframe. The fins basically self fillet internally and it uses less epoxy for less weight gain, the dams are made from either 1/8" or so balsa or basswood strips, trimmed flush with the CR's.

L2project8-2.jpgMiniFrenzyXLdd4.jpg
 
Excellent choice of rockets. The fin can pictured on the left is my 3" Frenzy XL using the same 5 CR technique you are planning to use ( the fifth CR is actually a thrust plate). On the right is the fincan for my Mini-FrenzyXL DD. Instead of using extra CR's and injected fillets like I did on the 3" I am using fin pockets. The fin pockets are the same technique I used on my L3 build and other rockets since. Basically the dams on each side of the fin are spaced to allow the size of fillet (internal fillet,that I want), then since they fit tight to the mmt and airframe both, the mmt is epoxied in to the airframe with the space between the dams lined up with the slots in the airframe. I can then use thicker epoxy that is inserted between the dams, then the fins are inserted and the excess epoxy that squeezes out is wiped off the exterior of the airframe. The fins basically self fillet internally and it uses less epoxy for less weight gain, the dams are made from either 1/8" or so balsa or basswood strips, trimmed flush with the CR's.

Huh. This is a really intriguing suggestion, thank you!. I'm already happy I started this thread! It means I can go to town on (and inspect) the internal fillets and don't have to drill holes in the body. First let me confirm my understanding. You are suggesting replacing this internal structure (black is fin, yellow is epoxy)
rect4193.png

with this (brown is the well material):
rect4197.png
My only concern: basswood (or something similar) will fail in tension much more easily than FG or epoxy, so doesn't this weaken the internal fillets? For example, the left fillet between the MMT and the fin offers support in compression when the root of the fin tries to move left and in tension when the root tries to move right. With the injection version, when in tension, the joint fails at the minimum failure strength of (MMT delaminates, epoxy->MMT fails, epoxy fails in tension, epoxy->fin fails, fin delaminates). When adding the well structure, it fails in tension at the minimum of all those as well as the well material (basswood) itself failing.

As such, shouldn't the well material be made out of FG itself? Basswood will fail in tension at a far lower force than anything else in the chain. Or maybe we don't worry about it because no matter which direction the force is in, there is always one side in compression and we rely on that to hold the fin.
 
You can make the dams out of any material you like, their only purpose is to contain the epoxy until it sets creating the fillet. If additional structure is sought they could be made from FG and filleted like your picture, I did it on my L3 project that way and have since determined additional strength was unnecessary. Maybe if I were going to fly your Frenzy on one of Loki's 54mm M motors the additional structure would be necessary along with tip to tip glassing the fins. The dams only need to be spaced off the root/side of the fin enough to provide the size fillet you want, the fillet is what provides the structure. Fillets between the dams and the mmt are not really necessary, I usually glue the dams in place with CA until the fin epoxy sets.
 
Last edited:
You can make the dams out of any material you like, their only purpose is to contain the epoxy until it sets creating the fillet. If additional structure is sought they could be made from FG and filleted like your picture, I did it on my L3 project that way and have since determined additional strength was unnecessary. Maybe if I were going to fly your Frenzy on one of Loki's 54mm M motors the additional structure would be necessary along with tip to tip glassing the fins. The dams only need to be spaced off the root/side of the fin enough to provide the size fillet you want, the fillet is what provides the structure. Fillets between the dams and the mmt are not really necessary, I usually glue the dams in place with CA until the fin epoxy sets.

Ah I see what you mean: the dams are just there to create a sealed well. I was thinking they would be nearly flush with the sides of the fin and therefore part of the structural support.
 
After eyeing this kit for some time I couldn't resist the good sale in November, and I'm ready to start building! I'd love to someday fly it on an M, but that won't be anytime soon since I want a lot more flights under my belt in the L2 range and MMMSC is limited to 10k feet.

I'm not making massive changes to the kit, but thus far I'm planning the following addendums/changes:

  • Use RocketPoxy everywhere (except motor retainer) and inject internal fillets. The injection process sounds like a mess to me, but so does doing them with a dowel between a 75mm mount and a 4" air frame with split fins and people seem to like the injection method so time to try it out.
  • Add two extra centering rings (making a total of 5) to make the dams for injection.
  • Aeropack 75mm retainer epoxied on with jb weld
  • 1" kevlar strap epoxied to two sides of the motor mount. The kit calls for attaching shock cord directly to motor, but I'll mostly be flying 54mm CTI which don't natively allow for that, and if I'm not adding a bulkhead might as well leave the motor charge in as a tertiary backup of the apogee split. I've been going back and forth for some time about the kevlar strap method vs adding a bulkhead, but (as with injecting) the kevlar strap method seems very popular so time to give it a shot
  • 2 eggtimer quantums in the avbay. TBD whether I add a physical switch, though I suppose it'd be required if I use this for L3
  • Eggfinder gps in the nose. I've already designed+built a twist-in enclosure to use the eggfinder in another rocket, so all I need is to cut a house-shaped hole in the nose cone bulkhead. More on this later.
  • Using the included ~30' TN shock cord for the booster -> avbay connection, and bought a 25' tubular kevlar cord to go from avbay -> nose. I'm on board with all the arguments against too-long cords, but I fly near trees and I want a shot at getting this back from one.
  • (tentatively) 60" fruity iris main and 15" fruity classic elliptical drogue. I've bought these but left them in the package in case the final weight is off from my estimates.
  • Using a fruity deployment bag with the main, probably with a 24" fruity elliptical for the pilot. I've used deployment bags before and I'm hooked.
  • I'm thinking about some sort of external mount for a couple of cameras. I don't mind the drag - this isn't a lean and mean rocket and I fly on the east coast, and I've got this crazy idea about doing either stereoscopic mounts for 3d or some sort of panorama.

Madcow lists 14lbs which I took as the dry kit weight, so I was assuming 17-18lbs or so fully built and loaded other than the motor. However, all the fiberglass parts and hardware supplied with the kit plus the aeropack and the parachutes only comes to ~13.25lbs, so even after I build, paint, build the avbay sled + electronics and add the gps, this may come out lighter than I expected. We'll see!

I'm glad to see your build thread for this! I too am soon to start construction of the same model, so I'll be checking your thead frequently.. and start my own thread once I get the go ahead for my preliminary construction package. It's funny..we have soo much in common.. I also have 2 more Eggtimer Quantum Kits to build, an Eggfinder TRS for the nose and have been troubling over the best way to do the internal fillets. Tight quarters!
I have lots of RocketPoxy and plan to use it for all bonds.. but I'm worried that it's too thick to get a good flow for the internal fillets. I saw another thread for a 3" FG Frenzy where Rocket Poxy fillets were inejected, but the epoxy didn't flow.. and thus there were only small segments of "fillet" below each injection hole. Here's a link. I think it's a current thread here.. it's a good thread with good pictures to evoke some thought.
I was also thinking of purchasing some Aeropoxy ES6209 high strength epoxy just for the internal fillets... I'm assuming that it will flow better.
As for fin installation order, I'm thinking there's no reason I can't install the forward fins first. Yes, they overlap the aft fins, but I think I was easily able to test fit install the aft fins after the forward fins were test fit in. If this is indeed possible, then I would install the 3 forward fins first, then come back and install internal fillets with a dowel.. with the aft fins out of the way. Just as you did, I also purchased 2 additional centering rings for use in fin attachment.
How well were the fin bevels done on your fins? The one's I got were a bit tapered at the edges and one or two had wider deeper bevels on one side vs. the other. I just bought a load of 60, 80 and 180 grit paper to use to fix the bevels on all the fins. I'm not looking forward to that, since it's a very manual process that is hard to get right.
I picked up ther Aeropack 54-75mm adapter.. as I plan to fly lots of K motors in this baby, but I will fly L's and eventually an M for my L3 certification sometime in the future. Thus, I'm going through channels to document the build per NAR L3 requirements etc.
 
Last edited:
I'm glad to see your build thread for this! I too am soon to start construction of the same model, so I'll be checking your thead frequently.. and start my own thread once I get the go ahead for my preliminary construction package. It's funny..we have soo much in common.. I also have 2 more Eggtimer Quantum Kits to build, an Eggfinder TRS for the nose and have been troubling over the best way to do the internal fillets. Tight quarters!
I have lots of RocketPoxy and plan to use it for all bonds.. but I'm worried that it's too thick to get a good flow for the internal fillets. I saw another thread for a 3" FG Frenzy where Rocket Poxy fillets were inejected, but the epoxy didn't flow.. and thus there were only small segments of "fillet" below each injection hole. Here's a link. I think it's a current thread here.. it's a good thread with good pictures to evoke some thought.
I was also thinking of purchasing some Aeropoxy ES6209 high strength epoxy just for the internal fillets... I'm assuming that it will flow better.
As for fin installation order, I'm thinking there's no reason I can't install the forward fins first. Yes, they overlap the aft fins, but I think I was easily able to test fit install the aft fins after the forward fins were test fit in. If this is indeed possible, then I would install the 3 forward fins first, then come back and install internal fillets with a dowel.. with the aft fins out of the way. Just as you did, I also purchased 2 additional centering rings for use in fin attachment.
How well were the fin bevels done on your fins? The one's I got were a bit tapered at the edges and one or two had wider deeper bevels on one side vs. the other. I just bought a load of 60, 80 and 180 grit paper to use to fix the bevels on all the fins. I'm not looking forward to that, since it's a very manual process that is hard to get right.
I picked up ther Aeropack 54-75mm adapter.. as I plan to fly lots of K motors in this baby, but I will fly L's and eventually an M for my L3 certification sometime in the future. Thus, I'm going through channels to document the build per NAR L3 requirements etc.

I agree, the internal fillets on my 3" Frenzy didn't flow well and while I think they'll be fine, I was already thinking that I wish I'd used fin pockets....
 
Last edited:
I'm planning to using fin pockets on my 3" FG build later this month. Can't wait to see this thing fly at MMMSC this summer.
 
I'm planning to using fin pockets on my 3" FG build later this month. Can't wait to see this thing fly at MMMSC this summer.
Yeah I'm sold on these pockets. Tentatively planning to cut some foam board to use for the dam. That way I can cut maybe 1mm too big and it will compress so I can fully fill the well without leakage. Tbd how wide I go. Id be skeptical of a square shaped fillet on a larger mmt -> body gap, but I think it will work great here. And no holes, leaks or trying to push rocket proxy through a syringe! I just need to not put the fin in too fast lest I get an epoxy fountain!
 
Last edited:
I'm glad to see your build thread for this! I too am soon to start construction of the same model, so I'll be checking your thead frequently.. and start my own thread once I get the go ahead for my preliminary construction package. It's funny..we have soo much in common.. I also have 2 more Eggtimer Quantum Kits to build, an Eggfinder TRS for the nose and have been troubling over the best way to do the internal fillets. Tight quarters!

Yup great minds think alike? I guess we'll see when it flies.... I'm a little worried that using 2 quantums is going to be unwieldy to arm but time will tell. Are you going ahead and putting in a physical switch to meet L3 requirements? I wasn't sure if I'd do that now or when and if I need it.

I have lots of RocketPoxy and plan to use it for all bonds.. but I'm worried that it's too thick to get a good flow for the internal fillets. I saw another thread for a 3" FG Frenzy where Rocket Poxy fillets were inejected, but the epoxy didn't flow.. and thus there were only small segments of "fillet" below each injection hole. Here's a link. I think it's a current thread here.. it's a good thread with good pictures to evoke some thought.
I was also thinking of purchasing some Aeropoxy ES6209 high strength epoxy just for the internal fillets... I'm assuming that it will flow better.
As for fin installation order, I'm thinking there's no reason I can't install the forward fins first. Yes, they overlap the aft fins, but I think I was easily able to test fit install the aft fins after the forward fins were test fit in. If this is indeed possible, then I would install the 3 forward fins first, then come back and install internal fillets with a dowel.. with the aft fins out of the way. Just as you did, I also purchased 2 additional centering rings for use in fin attachment.

Yup I've been following that thread. I'm probably going to screw in the nose cone coupler like he does but haven't committed. I like his fin guides but I already built one I'll talk about later that I might as well try. And his troubles with injection didn't make me excited about doing it myself :)

How well were the fin bevels done on your fins? The one's I got were a bit tapered at the edges and one or two had wider deeper bevels on one side vs. the other. I just bought a load of 60, 80 and 180 grit paper to use to fix the bevels on all the fins. I'm not looking forward to that, since it's a very manual process that is hard to get right.
The bevels are fine not great. Definitely some taper but didn't notice side to side asymmetry so I wasn't going to bother messing with them. I'm not sure I'd actually make them better without removing lots of material.

My bigger concern is that while the front fins are fine, one of the back fins is a little different size - the chord is about 1.5mm shorter and the height (how much it sticks out) is about 1.5mm larger than the other two. Interestingly, it's the same height as the forward fins while the other two are slightly shorter. I was debating asking Madcow if that was expected variance but I'll probably not bother. I'll sand down the taller fin and not worry about the chord being slightly shorter.

I picked up ther Aeropack 54-75mm adapter.. as I plan to fly lots of K motors in this baby, but I will fly L's and eventually an M for my L3 certification sometime in the future. Thus, I'm going through channels to document the build per NAR L3 requirements etc.
Yup I have that adapter. I'll probably start with a big J and then work up. Once I get a sense for how it flies I'll know if flying it with an M at my field is even conceivable. With any luck the cameras I add will make it draggy enough to keep it below 9k feet :) I've talked with an L3CC about this speculative L3 build in case that works out, but it's quite possible it won't. No sense in pushing it - even at 9k feet I'm likely to lose the rocket to a tree.
 
Do you guys really think the fin pockets are necessary? I can may be see it on some builds. But over 25 years I have never built a rocket with fin pockets. My idea is it just adds more weight and how much does it really help. Not trying to be a smart guy here but unless you think you can drop the booster with no chutes on black top and not have the fin pop off the MMT that's great but what keeps fin from breaking off above glue line. To each there own.
 
Do you guys really think the fin pockets are necessary? I can may be see it on some builds. But over 25 years I have never built a rocket with fin pockets. My idea is it just adds more weight and how much does it really help. Not trying to be a smart guy here but unless you think you can drop the booster with no chutes on black top and not have the fin pop off the MMT that's great but what keeps fin from breaking off above glue line. To each there own.

Couldn't give you a definitive answer, but the build manual for the kit strongly says that internal fillets are required to keep the fins on in flight given how short the ttw space is, and the pockets seem easier to me than using a dowel and pouring like the instructions say. This is a heavy heavy rocket already too (thick fg) so the pockets aren't going to change the weight all that meaningfully. Maybe a few ounces more on a 15 or 20lb loaded bird?

Its a good question though!
 
On to the building I've done so far. Nothing earth shattering here:

All the pieces laid out:

IMG_0057.jpg

1/8" vent hole in the payload bay:

IMG_0058.jpg

Marked up the motor mount tube with where the 5 centering rings will go as well as the kevlar strap:

IMG_0060.jpg

I stacked all 4 avbay bulkheads together and used my brother's drill press to drill a pair of holes 2" apart. Then epoxied them together and epoxied the switch ring to the avbay (after soap then lacquer thinner then sanding). I responded to the sticky but no responses so far - I wonder if it's OK to wet sand to prep the FG to have less dust?

IMG_0059.jpg

This all went pretty smoothly, though the bulkheads aren't quite aligned despite lining the pre-drilled holes up, so they will need a touch of sanding to fit on the bay. Not much, just a wee bit.

IMG_0062.jpg
 
Looks like you doing a great job on the build. I am just wondering if the pockets are worth the extra. I would be worried that the pockets would not be perfect straight and that the fin would not be able slid all the way down on the MMT if pockets were to tight.
 
Looks like you doing a great job on the build. I am just wondering if the pockets are worth the extra. I would be worried that the pockets would not be perfect straight and that the fin would not be able slid all the way down on the MMT if pockets were to tight.

Think of the pockets as nothing more than a method for auto-filleting the internals, thats why a very light material like balsa can be used it does nothing except keep the epoxy in contact with the airframe ( inside), motor tube, and the fin, how wide the dams are spaced off the fin is how wide the fillet will be. They are much quicker to make and less hassle than trying to fillet in the space between mmt and airframe, and they can use less epoxy than the injected fillet method.
 
Think of the pockets as nothing more than a method for auto-filleting the internals, thats why a very light material like balsa can be used it does nothing except keep the epoxy in contact with the airframe ( inside), motor tube, and the fin, how wide the dams are spaced off the fin is how wide the fillet will be. They are much quicker to make and less hassle than trying to fillet in the space between mmt and airframe, and they can use less epoxy than the injected fillet method.

Given that the epoxy dams have no role after the epoxy has set and adds no structural support, I wonder if there is a way that the dams could be removed after the epoxy sets enough to stop flowing. I'm imagining a a scenario where the rear centering ring is left off and the dam assembly is inserted from the aft end. Then, after filling the fin pockets with epoxy, and setting the fins, but before the epoxy is fully cured, pull the assembly back out. Assuming that the parts are 3D printed with PLA and your epoxy is acetone solvable, you could clean the parts and use them again for the next fin. Of course the timing would be critical but if you could make it work, you'd be able to make consistent internal "fillets", with no added parts. Plus, since you add the temporary fin pockets after the MMT is installed the alignment is flexible right up to the point that the epoxy is added. Just a thought...
 
Think of the pockets as nothing more than a method for auto-filleting the internals, thats why a very light material like balsa can be used it does nothing except keep the epoxy in contact with the airframe ( inside), motor tube, and the fin, how wide the dams are spaced off the fin is how wide the fillet will be. They are much quicker to make and less hassle than trying to fillet in the space between mmt and airframe, and they can use less epoxy than the injected fillet method.
Yup.

So the TTW space is a little less than 3/8". Any thoughts on how far from the fin the dam should be? 3/8" would make a square cross section, for what little it's worth.

The combined root chord of the two fins combined is about 13.5 inches though, so 3/8" wide means 13.5 * 3/8 * 3/8 = 1.9 cubic inches of epoxy per fillet on the two fins, or 6 * 1.9 = 11.4 cubic inches total. Rocketpoxy has a density of 1.5g/cm^3, so this would be 280g of epoxy or about 10z total. Ooof. Maybe I'll make them 1/4" wide unless there's something wrong with my math....
 
Yup.

So the TTW space is a little less than 3/8". Any thoughts on how far from the fin the dam should be? 3/8" would make a square cross section, for what little it's worth.

The combined root chord of the two fins combined is about 13.5 inches though, so 3/8" wide means 13.5 * 3/8 * 3/8 = 1.9 cubic inches of epoxy per fillet on the two fins, or 6 * 1.9 = 11.4 cubic inches total. Rocketpoxy has a density of 1.5g/cm^3, so this would be 280g of epoxy or about 10z total. Ooof. Maybe I'll make them 1/4" wide unless there's something wrong with my math....

Mine are usually a 1/4" or less wide on each side of the fin. And your math seems to be correct, and a 1/4" wide would cut the required epoxy by 1/3 roughly or 3.5oz for a total of 6.5oz.
 
Last edited:
I did the internal fillets on my Frenzy XL by just applying Aeropoxy with a dowel. It's a tight fit, messy job, takes a long time to do all those internal fillets and was a pain in the a$$, but it is doable.
 
OK on to the next item that may generate discussion :)

As I mentioned I'm using the popular glued-on-1"-kevlar method for a shock cord mount. I have a 6 foot piece from wildman of which I'll glue on 5.5" to each side of the motor mount (5.5" means I don't have to worry about the fins). The thing is, it seems to me that most of the 5.5" of bonding is wasted if the kevlar isn't under load while the epoxy hardens. As a I thought experiment, imagine if you bonded it on in a snake pattern like the following, where yellow is kevlar and brown is epoxy:

kevlarsnake.png

Obviously the only epoxy joint that matters in terms of how much force the strap can withstand is the one closest to the front. If/when that fails the next one is the only one under load and so forth. The spots in back add backup but not strength against the first bond failing. Now obviously we don't glue it on like this, but it seems to me that unless the kevlar is under load when the epoxy is applied, it's the same effect: the kevlar at the very beginning of the bond is held entirely by the epoxy under it rather than by the epoxy and the kevlar behind. Additionally, I imagine this will further decrease any movement of the kevlar fibers under load which would hasten the abrasive process.

So I epoxied just the very last 0.5" or so of the kevlar (so the part between 5" and 5.5" away from the front end of the MMT) and let it dry, using tape to hold the meat of it on. Tonight then I'll hang it under weight while I put epoxy under the rest.

IMG_0061.jpg

Now to be clear, I'm well aware that many very experienced folks (including legends like crazy jim) do the strap without the loading on much larger rockets and it works great, but part of the fun of this hobby for me is figuring out how/why things should work. I also realize that the much more likely failure mode of the kevlar strap is abrasion on the edge of the mmt or top of the rocket, or heat damage, so I'll be (1) making sure the epoxy ends before the last centering ring (0.5" from end of MMT) (2) sanding a rounded edge on the MMT tube (3) putting a nomex protector on the strap and (4) probably using tape around the top part so it can't chafe on the outer tube during descent.

Any thoughts, particularly about reasons this makes the attachment less strong rather than stronger?
 
OK on to the next item that may generate discussion :)

As I mentioned I'm using the popular glued-on-1"-kevlar method for a shock cord mount. I have a 6 foot piece from wildman of which I'll glue on 5.5" to each side of the motor mount (5.5" means I don't have to worry about the fins). The thing is, it seems to me that most of the 5.5" of bonding is wasted if the kevlar isn't under load while the epoxy hardens. As a I thought experiment, imagine if you bonded it on in a snake pattern like the following, where yellow is kevlar and brown is epoxy:

View attachment 309092

Obviously the only epoxy joint that matters in terms of how much force the strap can withstand is the one closest to the front. If/when that fails the next one is the only one under load and so forth. The spots in back add backup but not strength against the first bond failing. Now obviously we don't glue it on like this, but it seems to me that unless the kevlar is under load when the epoxy is applied, it's the same effect: the kevlar at the very beginning of the bond is held entirely by the epoxy under it rather than by the epoxy and the kevlar behind. Additionally, I imagine this will further decrease any movement of the kevlar fibers under load which would hasten the abrasive process.

So I epoxied just the very last 0.5" or so of the kevlar (so the part between 5" and 5.5" away from the front end of the MMT) and let it dry, using tape to hold the meat of it on. Tonight then I'll hang it under weight while I put epoxy under the rest.

View attachment 309091

Now to be clear, I'm well aware that many very experienced folks (including legends like crazy jim) do the strap without the loading on much larger rockets and it works great, but part of the fun of this hobby for me is figuring out how/why things should work. I also realize that the much more likely failure mode of the kevlar strap is abrasion on the edge of the mmt or top of the rocket, or heat damage, so I'll be (1) making sure the epoxy ends before the last centering ring (0.5" from end of MMT) (2) sanding a rounded edge on the MMT tube (3) putting a nomex protector on the strap and (4) probably using tape around the top part so it can't chafe on the outer tube during descent.

Any thoughts, particularly about reasons this makes the attachment less strong rather than stronger?

Wow, we think soo much alike... I'm detecting a mechanical engineering background.

I was also thinking about how the epoxy matrx Kevlar strap would transfer the load to the motor mount too. The one thing I had in my mind to look at was the modulus of elasticity for the fiberglass tube vs. the epoxy matrix Kevlar.. in longitudinal tension. I would assume that the Kevlar/epoxy would be much stiffer than the fiberglass, and thus achieve better load distribution. I just googled this which suggests the Kevlar is approximately 3 times stiffer than E glass. So, as the Kevlar is pulled.. the fiberglass stretches approximately 3 times more than the Kevlar/epoxy matrix... and thus the shear stress increases linearly along the bond length. I like your idea of preloading the Kevlar while the epoxy matrix cures is a good one... as it would minimize the amount of displacement caused by strain in the epoxy. I think what I'm getting from this is that the Kevlar/epoxy strip should be as stiff as reasonably possible. I think that adding some unidirectional carbon tow on top of the Kevlar/epoxy would help with the load distribution through the bonded length of strap too. Basically.. it might help even out of the shear stress through the length. I'm on a tight lunch schedule.. so I typed this quickly.. I hope it makes sense..
 
The only strap I've ever had break was a Kevlar strap, right at the edge of the epoxy, not because of abrasion and not because of heat, but what I felt was more like fatigue. I don't glue Kevlar anymore and whenever possible I use knotted tubular nylon and nomex sheaths.
 
The only strap I've ever had break was a Kevlar strap, right at the edge of the epoxy, not because of abrasion and not because of heat, but what I felt was more like fatigue. I don't glue Kevlar anymore and whenever possible I use knotted tubular nylon and nomex sheaths.
My understanding is that the epoxy edges at the edge of the bond act like a knife. Was the kevlar free to bend after the epoxy ended? In my case, I'm ending the epoxy before the foremost centering ring so that the kevlar is held flat mechanically by the ring and there will be little to no bending at the edge of the epoxy bond.

It is a worry though. I've also thought about whether it'd be better to have the knot at the top be asymmetrically placed so that only one side of the U loop is under load. This is half as strong, but it also means you'd have backup when the kevlar failed.
 
Wow, we think soo much alike... I'm detecting a mechanical engineering background.

Haha I'm flattered! Software Engineer by trade with a combined degree in Electrical Engineering and Computer Science. I have always loved physics and math and the limited parts of mech E I could get exposed to though, and have enough knowledge to at least be dangerous!

I was also thinking about how the epoxy matrx Kevlar strap would transfer the load to the motor mount too. The one thing I had in my mind to look at was the modulus of elasticity for the fiberglass tube vs. the epoxy matrix Kevlar.. in longitudinal tension. I would assume that the Kevlar/epoxy would be much stiffer than the fiberglass, and thus achieve better load distribution. I just googled this which suggests the Kevlar is approximately 3 times stiffer than E glass. So, as the Kevlar is pulled.. the fiberglass stretches approximately 3 times more than the Kevlar/epoxy matrix... and thus the shear stress increases linearly along the bond length. I like your idea of preloading the Kevlar while the epoxy matrix cures is a good one... as it would minimize the amount of displacement caused by strain in the epoxy. I think what I'm getting from this is that the Kevlar/epoxy strip should be as stiff as reasonably possible. I think that adding some unidirectional carbon tow on top of the Kevlar/epoxy would help with the load distribution through the bonded length of strap too. Basically.. it might help even out of the shear stress through the length. I'm on a tight lunch schedule.. so I typed this quickly.. I hope it makes sense..
Huh interesting. I wonder if those numbers are accurate for a woven strap laid in epoxy. Agreed about wanting it stiff - my thought experiment for this is a steel bar epoxied to a rubber surface vs a rubber band attached to the same surface: the steel bar intuitively will spread the load over the full bonded area much better. A carbon layover though seems like it might be excessive though. At that point why not just lay some sort of carbon fibre loop and epoxy that to the MMT?

As I mentioned earlier though, I think even just loading the strap will the epoxy sets (which I will do) is probably missing the forest for the trees - I'm pretty sure that if the strap fails it will be because it melts through, because it is worn away by a fiberglass edge, or because it is "cut" by the epoxy frontier far before the strap->mmt bond fails. And in my case I'm 99% sure I'll lose it to a tree before any of that lol.
 
Well the good news is
  1. I now have an attached strap
  2. I can tell you how *not* to do this :)

I set up a rig so I could hang a ~50lb toolbox from the strap and used a level to make sure it was all perfectly horizontal/vertical (the pre loading only helps if it is in the right direction!). I taped off the last 1" of the kevlar before the end of the MMT as well as along the sides of the strap to keep things neat. I used a pretty judicious amount of epoxy and hung it all up:
IMG_20170110_214121.jpg

And cue the entirely predictable (I foresaw it but didn't think it'd be a big deal) problem: under load the strap wants to be flat rather than conform to the tube, and it wants to do so with enough force that the unset epoxy wasn't enough to hold the edges of the strap against the tube. I actually suspect this is a major source of the rare times these straps fail: the outer edges of the strap take most of the load since it can't lay flat, which exacerbates any cutting/abrasive effects at the start of the epoxy joint.

After thinking things through, I decided that the foremost centering ring would hold the kevlar strap conformal to the tube, so I should just do the same with a compression wrap of tape. Fortunately I had been judicious with the epoxy so this didn't make a massive mess:
IMG_20170110_214653.jpg

I took the tape off this morning and looks pretty good: I have a nice clean end to the epoxy about 1" from the front end of the MMT, which should be far enough from the centering ring at 0.5" to prevent any substantial bending at the epoxy edge. As you can see I don't have epoxy slathered all over the place. This saved my butt from having a giant mess when I added compression wraps, but I'm not fully decided yet on whether I'm going to add epoxy on top. My instincts are not to: the kevlar strap is plenty capable of carrying the full load so might as well let it propagate the load down the bond without adding more epoxy in to potentially add cut points. I dunno though, it seems like the conventional wisdom is to butter it up pretty generously.

IMG_20170111_084003.jpgIMG_20170111_083954.jpg

I'm also undecided about whether I want to try and load this up with a few (or many :) ) hundreds of pounds to see what happens. The other good thing is that if/when I fly this on a 75mm I'll use a forged eye in the motor too as backup to these straps. That way there is no way this rocket will go out of sight and then lose the motor or entire booster section.

Anyway, if I do this again I'll put the front centering ring on before I lay the epoxy so that the strap is loaded exactly how it will be in the actual rocket and held conformally.
 
From caves.org:
Kevlar was one of the first high-strength fibers to be used in rope. It still offers high tensile strength and very low elongation, but has poor fatigue properties. The fibers inside the rope abrade each other, offering little indication of the reduced strength until the rope breaks. Kevlar has a very high melting point, 500 °C or 932 °F. Kevlar-core ropes are sold as escape lines for firefighters - to be used once and discarded. Manufacturers have had some success at solving the self-abrasion problem by combining Kevlar with Spectra.

https://caves.org/section/vertical/nh/49/cthsc/cthsc.html

So, I think that whether your strap conformed to the tube or not makes no substantial difference. Using the eyebolt in the forward closure is the right thing to do. Inspect the strap carefully before every launch.


Steve Shannon
 
Last edited:
From caves.org:
Kevlar was one of the first high-strength fibers to be used in rope. It still offers high tensile strength and very low elongation, but has poor fatigue properties. The fibers inside the rope abrade each other, offering little indication of the reduced strength until the rope breaks. Kevlar has a very high melting point, 500 °C or 932 °F. Kevlar-core ropes are sold as escape lines for firefighters - to be used once and discarded. Manufacturers have had some success at solving the self-abrasion problem by combining Kevlar with Spectra.

https://caves.org/section/vertical/nh/49/cthsc/cthsc.html

So, I think that whether your strap conformed to the tube or not makes no substantial difference. Using the eyebolt in the forward closure is the right thing to do. Inspect the strap carefully before every launch.


Steve Shannon
Huh this is great info, thanks for posting! I'll be sure to use the forward closure eyebolt when possible (including the L3 flight if I ever do it), though as mentioned before I'm mostly going to fly this on 54mm CTI which doesn't natively have that option.

Thoughts on whether to tie a knot at a slightly offset spot at the top? Using a knot like the butterfly seems like a good idea - that way if one side of the strap fails you haven't lost the rocket. Putting it in an offset spot means that you you only get the strength from one side, but the benefit is that the other side won't wear much (though it will still degrade from heat), so if/when the one side fails you still have a fresh strap on the other side to get the rocket down safely. And then you'd obviously retrofit a new strap before flying again. I'm thinking I may do it this way.
 
I would put the knot right at the top. Why deliberately put stress on one side or the other?
You're doing a nice build and I don't want you to think the Kevlar will snap easily. It'll take many flights and may never fail.
 
Well the good news is
  1. I now have an attached strap
  2. I can tell you how *not* to do this :)

I set up a rig so I could hang a ~50lb toolbox from the strap and used a level to make sure it was all perfectly horizontal/vertical (the pre loading only helps if it is in the right direction!). I taped off the last 1" of the kevlar before the end of the MMT as well as along the sides of the strap to keep things neat. I used a pretty judicious amount of epoxy and hung it all up:
View attachment 309157

And cue the entirely predictable (I foresaw it but didn't think it'd be a big deal) problem: under load the strap wants to be flat rather than conform to the tube, and it wants to do so with enough force that the unset epoxy wasn't enough to hold the edges of the strap against the tube. I actually suspect this is a major source of the rare times these straps fail: the outer edges of the strap take most of the load since it can't lay flat, which exacerbates any cutting/abrasive effects at the start of the epoxy joint.

After thinking things through, I decided that the foremost centering ring would hold the kevlar strap conformal to the tube, so I should just do the same with a compression wrap of tape. Fortunately I had been judicious with the epoxy so this didn't make a massive mess:
View attachment 309156

I took the tape off this morning and looks pretty good: I have a nice clean end to the epoxy about 1" from the front end of the MMT, which should be far enough from the centering ring at 0.5" to prevent any substantial bending at the epoxy edge. As you can see I don't have epoxy slathered all over the place. This saved my butt from having a giant mess when I added compression wraps, but I'm not fully decided yet on whether I'm going to add epoxy on top. My instincts are not to: the kevlar strap is plenty capable of carrying the full load so might as well let it propagate the load down the bond without adding more epoxy in to potentially add cut points. I dunno though, it seems like the conventional wisdom is to butter it up pretty generously.

View attachment 309154View attachment 309155

I'm also undecided about whether I want to try and load this up with a few (or many :) ) hundreds of pounds to see what happens. The other good thing is that if/when I fly this on a 75mm I'll use a forged eye in the motor too as backup to these straps. That way there is no way this rocket will go out of sight and then lose the motor or entire booster section.

Anyway, if I do this again I'll put the front centering ring on before I lay the epoxy so that the strap is loaded exactly how it will be in the actual rocket and held conformally.

Thanks for all the details and pictures. This is an excellent build thread with excellent discussion in just a few days already.

It is my understanding that we need to completely saturate the Kevlar strap through each of the bond lengths to the motor tube. This epoxy is the "matrix" for the Kevlar fiber reinforcement.. together forming the composite material. To get the saturation, a thinner epoxy can be used, or perhaps heat will initially thin the epoxy sufficiently to get improved saturation/flow into the Kevlar weave.
From Wikipedia:
"Composites are made up of individual materials referred to as constituent materials. There are two main categories of constituent materials: matrix (binder) and reinforcement. At least one portion of each type is required. The matrix material surrounds and supports the reinforcement materials by maintaining their relative positions. The reinforcements impart their special mechanical and physical properties to enhance the matrix properties. A synergism produces material properties unavailable from the individual constituent materials, while the wide variety of matrix and strengthening materials allows the designer of the product or structure to choose an optimum combination."
Without completely saturating the Kevlar, the load distribution will be nearly all at the top of the bonded length with very little load distribution down the bonded length. Only the Kevlar fibers that are inside the epoxy you used to bond the underside of the strap will be in a matrix that will allow the Kevlar fiber to distribute load.

I like the idea of a section of the Kevlar strap just above the epoxy bond being "strain releaved" using an outer sleeve and/or a wrap around the motor tube. Generally, however, the tensile load direction in the Kevlar strap should be ~straight up the booster tube.. there *should* be little force acting in any other direction.
 
Back
Top