December 20, 2010 at 12:43 am #41157
Warren B. MusselmanModerator
There have been a few threads recently that got me to thinking that perhaps I should put together an article on my L3 project. This project was an all-composite build with a number of features that I hadn’t seen anyone do – at least up until then and I thought that going over the details along with posting some pictures would be helpful to some of the folks who are interested in composite construction and vacuum-bagging.
Essentially, I started with a Performance Rocketry Competitor 5. No reason I chose that kit except that Tim from Wildman came to one of our launches and he gave me a really, really, REALLY good price on the kit – like $100 cheaper than the components would cost me to buy direct from PR. The kit is a classic 3FNC bird with an ogive nosecone and a 98mm motor mount. I sat on the kit for a year or two before I actually started the build and spent a lot of time thinking about just what I wanted the bird to accomplish and what kind of motor flexibility I wanted it to have. In the end I wanted the project to be a fun sport rocket able to fly big 98mm motors, further refine my composite construction techniques, provide a platform for testing various rocketry electronics projects I had in mind and finally to be my L3 cert bird. I also wanted to build the bird strong enough to survive in case of a failure to deploy the main – something that had happened to me a time or two on the journey to my L2 cert.
Along the way I must have spent nearly 100 hours in Rocksim fiddling with various design variations. In the end, I decided that I wanted to be able to fly up to the largest motors I could fit in the airframe. The fiberglass motor mount tube wasn’t nearly big enough at a mere 18” so I replaced it with a 34” piece of PML phenolic. Phenolic is a much better insulator than fiberglass thereby limiting heat soak problems and the increased length would allow me to fly 5 grain N motors and still have enough space in the 48” booster airframe to stuff a drogue and shock cord. I could probably fit a 6 grain N motor into the airframe if I dispensed with the drogue. I also modified the straight delta fins to clipped delta planform, because of slightly better aerodynamics and primarily because I like the look better. I also decided to cut a 1 ¼” section of airframe off to act as a spacer on the coupler and provide a place to mount my electronics switches.
At the time, I had access to a precision machine shop and was eager to learn to use the lathe and 5-axis NC mill we had, so I decided that instead of mounting my Aeropac retainer to the kits’ stock G10 ring, I would design and machine an aluminum thrust-ring for the rear of the bird. I also made some additional aluminum centering rings just for the practice and some stainless steel rail buttons since I wasn’t thrilled with the commercial rail buttons I had found. The thrust ring was machined to take the motor mount tube on the inside with a step and to fit flush with the outside of the airframe with a step to fit inside the airframe tube.
Thrust ring forward face
Thrust ring rear face
Thrust ring with Aeropac retainer
Stainless Rail Buttons
This design has through-the-wall fins that are attached to the motor mount tube. I also wanted to do a tip to tip layup that would use multiple types of fabric and that would taper from root to tip in order to preclude resonant vibration in the fins and fin “flapping” in high-G flights. This concept is too complicated to go into here, but I believe there are other posts discussing it here in the Composite Construction Forum. Also, in order to provide the highest strength, I wanted the forward and rear ends of the fins “trapped” between centering rings to provide greater strength in the event the rocket impacted on a fin during a bad recovery.
To start the build, I began by preparing the fins and the phenolic motor mount tubing. The tubing was thoroughly sanded with 80 grit sandpaper as was the pre-cut and beveled fins that came with the kit. I swapped out the blade on my table saw with a masonry cutoff wheel and then cut the full delta fins off about an 1 ½” to produce the clipped-delta planform I desired. I then made two cardboard alignment templates so that the fins could be glued square to the tubing. The fins were then glued to the motor mount tube with JB Weld epoxy. If you want to be anal, I would use the Cotronics high temp epoxy here instead. The alignment jigs are used to hold the fins in alignment and tight against the tubing with bungie cords to ensure a solid root joint.
Cardboard fin alignment template
Fins epoxied to motor mount tube using alignment jigs and bungie cords
Following this, I installed the centering rings at the front and rear of the fins and also used JB Weld. Finally, I filleted all joints with epoxy thickened with Kevlar pulp, chopped fiberglass and some fumed silica. The fillets provide a smooth curve for the fabric to lay across, thereby making the layup easier and more importantly by relieving the strain that is induced by laying the fabric across a tight 90 degree angle. The fabric is considerably weakened by such an abrupt angle whereas the smooth curve allows the fabric to remain near 100% of full strength.
Centering rings installed above and below fins – NOTE fillets on fin and centering ring joints all around
The next step was to do the initial layers of tip-to-tip fabric layup. For impact strength, I used a first layer of 5.7 oz. standard weave Kevlar that extended across the fin/motor tube joint. On top of this I laid up a larger layer of unidirectional Kevlar perpendicular to the airframe. Finally, I laid up a still-larger layer of 6oz S-glass fiberglass. These layers were each individually vacuum-bagged using non-stretch mylar v-bag material, peel-ply and breather fabric to compress them as tightly as possible to the fins and motor tube and to suck them into the fin/tube joint as well as to remove as much excess epoxy as possible.
First layer Kevlar – partial tip to tip across fin joints
Vacuum bag assembly around layer 1
Following the tip to tip layup across the motor mount tube, I laid up two layers of ring shaped glass across the forward side of the ring/tube joint and then installed the forward motor mount centering ring and the two 3/8” u-bolts that would be used for the shock cord mounts. This ring was a composite of the kit G10 ring and a machined aluminum ring I made on the lathe. As before, all parts were thoroughly sanded with 80 grit wet/dry paper, laminated together and then epoxied to the motor mount tube about 3/8” back from the forward end with JB Weld. Unfortunately, a 5” rocket with a 4” motor mount didn’t provide enough space for the 3/8” nuts on the u-bolts so I had to turn down the nuts on a lathe and knurl them to allow them to be tightened. After the u-bolts were installed, I filleted the centering ring to motor tube joint as were the fins and fin centering rings.
Forward centering ring showing knurled nuts and composite ring
After installing the u-bolts I wrapped the upper half of the motor mount tube with a layer of 10oz E-glass fiberglass to strengthen the motor tube. This layer lapped both the forward fin centering ring and the rear face of the forward ring.
Completed Motor Mount and Fin assembly – note glass wrap on tube
At this point, the fin/motor mount assembly was complete and it was almost time to install it into the airframe. First I filled the layup on all the fins with epoxy thickened with fumed silica (now I would use Aeropoxy Lite for the filler) and sanded things smooth. The fin slots in the airframe had to be enlarged to fit the thicker, tapered fins and the inside to the airframe had to be sanded to bond well with the epoxy on the inside of the fin slots and where the rings would contact the airframe. First I marked the airframe where the rings would contact so I used a Dremel tool with a 40 grit flapper wheel to do this. I also had to mount the T-nut that would support the rear rail-button so I trimmed down the T-nut on a grinding wheel and knurled the shaft on the lathe to provide something for the epoxy to bond to. Finally, I used a cut-off wheel on the Dremel to cut through the fin slots at the rear of the airframe to allow it to be spread apart to fit over the fins.
Inside of airframe showing roughed up areas
Airframe showing slots split and rail button mount
Next, I painted the inside of the airframe where the rings would contact with epoxy thickened with chopped fiberglass and installed the fin/motor tube assembly. As the assembly was inserted, I painted addition thickened epoxy on the outside of the rings and also added addition epoxy on the forward faces of the rings so that it would run into the joint from gravity after it was installed. I compressed the tubing around the centering rings with very tight bungie cords.
Painting epoxy on the rings prior to installation
Mount/Fin Can installed in airframe and bungie compression
Following this I used a syringe to fill the space around and between the forward end of the motor tube and the airframe and then filleted the fin joints on the outside to seal the relatively rough-edged fin slots. I also drilled ¼” holes between the fins next to the rings so that I could inject epoxy to fillet the inner joints after the motor mount was installed. Unfortunately I don’t have clear pics of this as I didn’t have an assistant that day. After injecting the epoxy, which I did in several sessions, I angled the assembly to allow gravity to draw the epoxy into the joint I was trying to fillet that session. As I recall, it took a total of 5 sessions to complete the internal fillets on all the fins and rings. These holes were then used to fill all the spaces between the rings by injecting 2-part foam. This also took a couple sessions to complete.
Thickened epoxy puddled around top of motor mount tube
Fin joint sealing fillets and tape
Fabric filled, fin fillets applied
After the inner-filleting and foam was completed, I first filled the drilled holes with thickened epoxy and red auto body filler and sanded them smooth.
Inner fillets and 2-part foam complete
Then I began the final layups of tip-to-tip. This was done with alternating layers of 5.6oz standard weave carbon fiber and 6oz E-glass with each layer slightly larger than the previous layer. In the end there was a total of 8 layers vacuum-bagged down. In between each layup, I again filled the irregularities with epoxy thickened with fumed silica and sanded them smooth. Each layer was larger than the previous layer to continue to increase the fin taper from root to tip – the last 2 layers were full size and extended about a 1/2″ past the edges of the fins. I don’t show the layup of each layer, but the basic sequence is show in the next few pictures. Essentially the process is to wet out a piece of fabric and apply it, then apply a layer of peel ply, then a layer of breather fabric and finally to assemble the bag around the the whole fin can and pull vacuum. This is the critical part of building a unitary fin can.
Beginning the layup
Adding the peel-ply
Bagging the layup – constructing the bag around the fin can is the hard part of the whole job
This process is repeated for each layer of the layup, or once you get the hang of it, you can do 2 or even 3 layers at once although you’ll need to add an additional layer of breather and pay particular attention to wrinkles.
Final layer is on – note epoxy wicking into the breather layer
After the bag, breather and peel-ply is stripped. The edges still need to be trimmed with a sharp knife
When I did this project, I v-bagged each layer individually and did some fill-work in between layers so that I didn’t get wrinkles. If you do get wrinkles, you’ll have to pull everything off quickly before the epoxy sets and wash it all down with acetone to clean up or your layup is ruined. As I’ve gotten more experience, I’ve moved to doing 2 or even 3 layers at a time. THE keys to getting this process right is to have ALL your materials pre-cut and staged as well one or more assistants to hand you fabric pieces, hold things in place and to help you seal the edges of the bag in a quick and efficient manner. Once you mix the epoxy it is far too late to cut pieces or rummage around for tape. Also, use the longest pot-life epoxy you can get so that you have time to do the layup before the epoxy kicks off.
As for the vacuum bagging process, I use West Systems epoxy for layups and I pull a vacuum until the epoxy has set green. If you use an epoxy that has to be cured at elevated temperatures, you have more time to get things right although you’ll have to use mylar bagging material that can be placed in a curing oven. I usually leave a scrap of fabric wet out at the same time as my layup on the bench so I can test how far the cure has gone so I know when I can strip the bag and peel-ply. Even with peel-ply leaving a rough finish for further bonding layers, I consider it critical that you don’t wait for a full cure before applying the next layer – you want the epoxy to still be green (under 24 hours) when you apply the next layer so that you can get a full chemical bond (cross-linking) for the maximum possible layup strength between the layers.
If you do use an epoxy that requires a heat cure, there are some issues with managing the vacuum hose. A rotisserie type curing oven is also impossible due to the hose so your curing oven will have to be built to avoid hot spots. After screwing around with that process, I went back to West Systems and other self-curing epoxies like Pro-set as the hassle-factor was just too great without a commercial autoclave. On smaller projects, I will do an elevated post-cure by sticking it my wife’s oven on Warm with the door open or on the dashboard of the car in full sun on a hot day. However, with West Systems, going beyond 125 degrees won’t buy you anything and exceeding that temperature will damage the crosslinking in the epoxy thereby reducing final strength.
I want to thank John “3 Dogs” Wilke, Jon Skuba, Dale Netherton, Eric “Elvis” Parsons, Kyle Parsons and Jeffery Joe Hinton for helping me develop these techniques during the build of The UprOar project. I especially want to thank Joe Hinton for being a tireless and incredibly giving and helpful friend who has helped me with a number of layup projects including this L3 project. As a club, we are truly blessed with some of the smartest and most helpful folks I have ever met. Thanks to you all.
LIFTOFF – First flight
PS: I forgot to mention, each fin has a total of 16 layers of fabric at the root and 4 layers at the edges. Layers 1, 2 and 3 go across the motor mount tube.
The layer pattern is:
1 – 1/4 span standard weave 5.7oz Kevlar
2 – 1/3 span unidirectional 4.5 oz Kevlar – fibers 90 deg to airframe (across span)
3 – 1/2 span standard weave 6 oz E-glass fiberglass
[Install in airframe]
4 – 1/4 span standard weave 5.6 oz Carbon fiber
5 – 1/2 span standard weave 5.6 oz Carbon fiber, weave biased 45 deg
6 – 3/4 span standard weave 6 oz E-glass fiberglass
7 – Full span standard weave 5.6 oz Carbon fiber
8 – Full span standard weave 5.6 oz Carbon fiber
One other thing, I usually top everything with a layer of 6 oz E-glass as a sanding veil – something I can sand through during the finishing process without compromising strength. I didn’t on this – probably due to laziness and because I had 2 layers of carbon on each side covering the entire fin. As it was, these fins were serious overkill – a 215# friend of mine could stand on the horizontal airframe with the fins tips on the floor with no detectable flex. (Fin cores are 1/4″ G10 plate – solid enough alone for the L3 flight. I did this to try the technique before doing a set of fins made of Nomex 1/8″ honeycomb for another project – that project still hasn’t come together, but lets just say it is a 4″ minimum diameter bird.)December 20, 2010 at 3:27 am #53639
Thanks alot for that great post Warren! Very interesting to see what can be done with a vacuum bagger. What did you use to attach your shock cords to the u-bolts? It looks to tight to use a quicklink without it contacting the main airframe under load.December 20, 2010 at 3:43 am #53640
Warren B. MusselmanModerator
Those U-bolts are there purely for backup in case the eyebolt in the motor closure fails or unscrews itself. I use 1/4″ quicklinks – 1 per U-bolt and I have a short loop of sewn kevlar set up in a load balancing arrangement (Take the loop and make a figure 8 shape with it – clip at the ‘X’ for the single point attachment. The 1/4″ quicklinks clear the side of the airframe though if I had it to do over again, I’d bend the U-bolts in towards the center a 1/2″ or so – that would clear even 1/2″ quicklinks and probably still allows space for a motor to fit past them if I use a 5 or 6 grain case although that wouldn’t leave any space for quicklinks at all. In that case I’d use knotted 1/8″ Kevlar as the fail-safe loop.
All that said, the primary load is carried by the Eyebolt in the forward closure of the motor – these U-bolts are only there as a backup and there is substantial slack in the loop when the shock cord goes tight to the motor eyebolt.
If that doesn’t make sense, let me know and I’ll take a picture.
You could use old D-shaped carabiners and clear – clearance at the u-bolts is a hair over 3/8″.
WDecember 20, 2010 at 5:50 am #53641
Warren, very nice article on a well built, beautiful bird. Thanks for taking the time and showing some of your tips.December 30, 2010 at 10:30 am #53642
Wow – impressive build, and I hadn’t realized that you had that much fabric on the fins. I should drag race you sometime 😀
You must be logged in to reply to this topic.