75mm Composite Minimum Diameter Build

[LandonC]

@watheyak,
The current OpenRocket sim puts our N5800 around Mach 3.5, but we're assuming that's a healthy over-projection considering RasAero and RockSim place it lower right around 2.8-9. It may seem overkill to use such a big motor, but it's the exact one we already use as the main motor in our Spaceport two-stage rocket, so it's familiar and we'd like to do more testing on it anyway.

@robopup,
I'm working on a calculator to give us stagnation temperature using sim data. We've found OR isn't reliable with Mach projections, so I'm trying to figure out all the math behind it so we have something solid to go off of. I want to be able to confidently say how long our rocket will be spending in each Mach region because if we only break Mach xyz for a split second and the duration of the flight is in xyz region, we can have a better idea of the stresses the rocket will experience. Also still working on sims to calculate heating.

I'll talk about this in my own build thread once we have a pretty product to show for it, but right now we're figuring out our NC. We're obviously going to have an aluminum tip, but we're also going to be putting temperature probes throughout the NC and tip so that we can compare to sim data. This is what we want to know for our larger two-stage rocket, although heating effects aren't nearly as big of a concern with that rocket.

Here's an idea I want to run past some people. If we're concerned about heating in the NC, would it be beneficial to have a layer of kevlar built into the layup? The entire NC was going to be fiberglass to allow for our eBay to function properly.

 

[waltr]

@watheyak,
Here's an idea I want to run past some people. If we're concerned about heating in the NC, would it be beneficial to have a layer of kevlar built into the layup? The entire NC was going to be fiberglass to allow for our eBay to function properly.

I don't think it is the cloth used that is affected by the heat. It is the epoxy matrix that one needs to be concerned about.

FG for ebay? Is that due to have an RF transmitter in the ebay?
Both FG & Kevlar are RF transparent whereas CF is not.

 

[watheyak]

@watheyak,
The current OpenRocket sim puts our N5800 around Mach 3.5, but we're assuming that's a healthy over-projection considering RasAero and RockSim place it lower right around 2.8-9. It may seem overkill to use such a big motor, but it's the exact one we already use as the main motor in our Spaceport two-stage rocket, so it's familiar and we'd like to do more testing on it anyway.

@robopup,
I'm working on a calculator to give us stagnation temperature using sim data. We've found OR isn't reliable with Mach projections, so I'm trying to figure out all the math behind it so we have something solid to go off of. I want to be able to confidently say how long our rocket will be spending in each Mach region because if we only break Mach xyz for a split second and the duration of the flight is in xyz region, we can have a better idea of the stresses the rocket will experience. Also still working on sims to calculate heating.

I'll talk about this in my own build thread once we have a pretty product to show for it, but right now we're figuring out our NC. We're obviously going to have an aluminum tip, but we're also going to be putting temperature probes throughout the NC and tip so that we can compare to sim data. This is what we want to know for our larger two-stage rocket, although heating effects aren't nearly as big of a concern with that rocket.

Here's an idea I want to run past some people. If we're concerned about heating in the NC, would it be beneficial to have a layer of kevlar built into the layup? The entire NC was going to be fiberglass to allow for our eBay to function properly.

With an N5800 you most certainly need to do something about heat. Whether that's high temp epoxy and structural adhesives, ablative, or both. The added complication of making sure all your other components, including vacuum bagging materials and molds, can handle the high post-cure temps adds significant cost and complexity. Oh yeah, dont forget an oven big enough for the largest component.

Kevlar vs fiberglass won't make much difference when it comes to heat resistance. It's the laminating resin that's the problem. I only did Kevlar in my L record project nosecone because I thought it would look cool. It's also a massive pain. Hard to cut, hard to drape, and you can't sand it once the part is finished.

This is a subject I've put a lot of time, sweat and money into. I'm not trying to discourage you, but just be aware that when high temperatures and composites meet, the work, and expense increase dramatically.

I'd read up on other minimum diameter N5800 projects here on the forum to get a feel for what worked and what didn't.

 

[robopup]

It's an N5800 now?! I appreciate the ambition (and I hate being that buzzkill guy now), but I can't imagine attempting that for my first minimum diameter flight (I believe?). Have you considered something smaller, say a single stage 54mm, as a first go around? Trust me, there's plenty of complexity there already; just see watheyak's L2050 thread.

And Waltr is correct, for your purposes it's the matrix you want to be worrying about w/r/t failure due to heat, not the fiber.

 

[robopup]

I should add, I'd take another look at your sims between Rasaero and OR. I wouldn't expect that much of a difference in max speed between the two. You'll notice my original sims for this project are within a couple percent for altitude (and similar for velocity if I remember correctly). For our purposes here, they all implement the same equations of motion. Assuming you have exactly the same masses, dimensions, atmosphere model, angle on pad, motor data etc, the biggest difference you're going to see is the drag model (and AFAIK Rasero's drag model, at least at these speeds, is more sophisticated).

 

[LandonC]

I don't think it is the cloth used that is affected by the heat. It is the epoxy matrix that one needs to be concerned about.

FG for ebay? Is that due to have an RF transmitter in the ebay?
Both FG & Kevlar are RF transparent whereas CF is not.

Yes, we will have an eBay with transmitters in the NC. This is the resin I found on Fibreglast that we'll probably end up using. Haven't decided for sure. https://www.fibreglast.com/product/System-4600-Epoxy-Resin

It's an N5800 now?! I appreciate the ambition (and I hate being that buzzkill guy now), but I can't imagine attempting that for my first minimum diameter flight (I believe?). Have you considered something smaller, say a single stage 54mm, as a first go around? Trust me, there's plenty of complexity there already; just see watheyak's L2050 thread.

And Waltr is correct, for your purposes it's the matrix you want to be worrying about w/r/t failure due to heat, not the fiber.

I don't have a say in the motor type. Senior members decided that gave us the challenge that we wanted. I've made smaller L1-sized min diameters, and other members have made slightly larger rockets, but this is obviously the largest project we've done. I appreciate all of the feedback and I recognize that we have our work cut out for us. But I live for the challenge. It'll be a fun next few months.

As to the sims, I double-checked and there were some slight parameters I forgot to set the same between the two. Thank you for pointing that out.

 

[LandonC]

With an N5800 you most certainly need to do something about heat. Whether that's high temp epoxy and structural adhesives, ablative, or both. The added complication of making sure all your other components, including vacuum bagging materials and molds, can handle the high post-cure temps adds significant cost and complexity. Oh yeah, dont forget an oven big enough for the largest component.

Kevlar vs fiberglass won't make much difference when it comes to heat resistance. It's the laminating resin that's the problem. I only did Kevlar in my L record project nosecone because I thought it would look cool. It's also a massive pain. Hard to cut, hard to drape, and you can't sand it once the part is finished.

This is a subject I've put a lot of time, sweat and money into. I'm not trying to discourage you, but just be aware that when high temperatures and composites meet, the work, and expense increase dramatically.

I'd read up on other minimum diameter N5800 projects here on the forum to get a feel for what worked and what didn't.

Good to know, thank you! I'll post an update in a bit on our progress.

 

[LandonC]

Sorry for clogging your thread with our issues...

We can switch over to mine in the future.

 

[robopup]

Sorry for clogging your thread with our issues...

We can switch over to mine in the future.

All good! I'll keep an eye on your thread...

 

[G_T]

With an N5800 you most certainly need to do something about heat. Whether that's high temp epoxy and structural adhesives, ablative, or both. The added complication of making sure all your other components, including vacuum bagging materials and molds, can handle the high post-cure temps adds significant cost and complexity. Oh yeah, dont forget an oven big enough for the largest component.

Kevlar vs fiberglass won't make much difference when it comes to heat resistance. It's the laminating resin that's the problem. I only did Kevlar in my L record project nosecone because I thought it would look cool. It's also a massive pain. Hard to cut, hard to drape, and you can't sand it once the part is finished.

This is a subject I've put a lot of time, sweat and money into. I'm not trying to discourage you, but just be aware that when high temperatures and composites meet, the work, and expense increase dramatically.

I'd read up on other minimum diameter N5800 projects here on the forum to get a feel for what worked and what didn't.

Just thought I'd mention about Kevlar since I've done a fair bit with it in the past:

It drapes fairly well if used on a bias. But fabric choice does matter. I like twill weave for instance.

It cuts like paper if using proper or properly prepaired shears. Get some cheap Fisker's or similar. Take it to a coarse belt or disk sander, and make a pass at 90 degrees to the flat contact side of the blades. Kevlar cuts very poorly by cutting, but shears well. The resulting searated rough surface prevents the Kevlar fibers from sliding away from the cut. The result really does cut like paper. You'll be amazed.

Kevlar sands well enough if you wet sand it with very good quality sandpaper and the epoxy matrix is fully cured. The sandpaper will work best if it is a bit open in grit structure, to help prevent any tendency to clog. Just mentioning this in case you have a reason to do it again.

All that said, I'd not be using Kevlar in a nosecone generally. I would use Kevlar under these conditions:

1) Protect a human occupant from carbon fiber needle death in event of an accident in a vehicle. Kevlar holds things together, mostly, sort of like a bag. Carbon fiber however makes nice razer sharp points and edges when it fails.

2) When vibration dampening is required. Kevlar absorbs a fair bit of energy each oscillation. Carbon fiber returns nearly all the energy each oscillation. So in a fin layup there might be some cause to use some Kevlar to slightly delay when flutter blows it apart. In any event, Kevlar is not the only fabric with this property nowdays and may not be the best one to use for your application.

3) You got it cheap.

4) You just want the golden look. But I recommend alternating layers with S-glass in that case. Kevlar stinks for compression strength and likes to crinkle/crunch. Kevlar also can delaminate between layers a bit more easily than other fabrics, for some reason.

5) Actually, for a nosecone I'd just use S-glass or possibly basalt fabric if you have a supply of good material. Once upon a time you used to be able to get good/interesting surplus fabric out of Russia. Decades ago actually. It won't be as consistent as S-glass but that might not matter. The properties of these two fabrics are quite close, both will bond well, both can take high temperature, both are good for compression and tension, etc. I never worked with basalt fabric.

--- For those that don't know, S-glass is structural glass. E-glass is mostly what you find being sold to hobbyists (E for electrical; originally for insulation). The difference is considerable.

--- Also thought I'd mention that I'd expect to have a much higher solids loading in a mold tooling coat than the numbers I've seen in this thread. Otherwise your properties are still essentially not much different than just the properties of the epoxy. You want enough in there that the properties shift a long ways towards the properties of the filler materials. You'd have to test, but I'd be aiming for well over 50% fillers by mass (except for low density fillers such as glass or phenolic microballoons), perhaps even up around 80%. I don't recall what numbers I used to use. It will certainly depend on the density of your fillers. You'll need to progressively mix in the fillers.

Gerald

 

[robopup]

--- Also thought I'd mention that I'd expect to have a much higher solids loading in a mold tooling coat than the numbers I've seen in this thread. Otherwise your properties are still essentially not much different than just the properties of the epoxy. You want enough in there that the properties shift a long ways towards the properties of the filler materials. You'd have to test, but I'd be aiming for well over 50% fillers by mass (except for low density fillers such as glass or phenolic microballoons), perhaps even up around 80%. I don't recall what numbers I used to use. It will certainly depend on the density of your fillers. You'll need to progressively mix in the fillers.

Gerald

Hey Gerald, when you go "much higher solids" on your tooling coat, how do you pop all the bubbles when you apply it to your plug? I stole my numbers from wyowindworks on various rcgroups threads which I suspect you're familiar with. With that type of consistency I can pop bubbles with a toothpick or propane torch. Beyond that I was never successful in my small tests. Also, note even fairly low volume fractions can have a significant impact of homogenized material properties of the matrix+particle. For example, here's a very quick plot I put together of stiffness versus volume fraction for standard modulus epoxy and S-glass particles:

Unit cell that I'm homogenizing:


Plot:



Another attempt is for the most part on the backburner right now as I make things even more complicated here. I did lay up another nosecone, print another ebay and sew another parachute for a new try. I have three 6 grain 75mm motors ready to go for my two stage project. If, in the unlikely event that I go 2/2 in my two stage attempt, the third motor will be slated for this project. Likely I'll just use one of my 75mm booster fincans which is not optimized for this flight, but will work.

On some holiday travel last Thanksgiving, I happened to fly right over Blackrock. I kept my eyes peeled for some red or orange out the window, but at 35k ft... good luck! Was still awesome to fly over though.

 

[G_T]

I put the tooling coats on thin, and yes use a heat source to float the bubbles. Then pop with a needle. But part of it is to mix slowly rather than vigorously to minimize how much air is in there in the first place, use a very watery laminating resin such as MGS, and it helps if you vacuum the mix before applying. Heat gun would be better than torch for raising bubbles. Torch can locally overheat and rapid-cure, making the result more brittle, more stressed, and weaker. Note you can go bimodal on your particle additives to help boost fluidity at higher solids loading.

I may end up making a nosecone mold for a 114mm. I have a heavy-weight 114mm nosecone for a blackrock project, but it is overkill (overly heavy). I think I can achieve the strength needed at about half the weight. I'd probably make the mold rather like I've made molds for sailplane fuselages, and then bladder mold. Though not to as high a pressure as I usually use! If I did that, with the large diameter, I'd have to make the mold walls thicker than I'd want to keep it from becoming a bomb. I don't need another heavy mold to store.

Gerald

PS - If people are interested in expensive 114mm nosecones let me know. That may encourage me to make the molds. 114mm is a nice size for those who do not use commercial hardware. Nice for good O up to small P motors.