L3 Winter Build Thread - 3/4 Scale PAC-3 Patriot

The Rocketry Forum

Help Support The Rocketry Forum:

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

mccordmw

Well-Known Member
TRF Supporter
Joined
Jan 18, 2016
Messages
1,051
Reaction score
72
Location
St. Louis, MO
I've been working on the specs, designs, and calculations on this for quite a while. The Patriot is a beautiful design, and I wanted to use that for my L3 build. I love the look of the MIM104F next generation version, so here it begins. I've got a ton of calculations to post before I get to the build portion.

So to kick it off, here's the summary pic and first release version of the ORK file.

2016-11-29 15_56_01-Rocket (8 inch scratch MEADS PAC-3 L3.ork)-3d.jpg

2016-11-29 15_55_32-Rocket (8 inch scratch MEADS PAC-3 L3.ork).jpg

I want to use is as a workhorse for (much much later) EX 75-98 motors once I've mastered formulations in the 24-29mm range. So this was built with a 98mm motor mount and an Aero Pack 98>75 adapter in the design to house my Loki 76mm M1969 sparky. At a 6.8:1 TTW ratio, it's a bit close. If I get too heavy, I'll bump up to an M3400 at a nice 10:1 ratio.

Stuff I want to cover before the build portion:
  1. the airframe and fiberglassing strategy
  2. internal structural load-bearing specs
  3. fin flutter analyses to calculate safe speeds
  4. fin strength to check safe AoA parameters
  5. electronics
  6. deployment scheme

--- the build ---

Then:


  1. chute calculations
  2. weight and motors
  3. flight profiles

I think that covers most of it. Going to take a while... I guess that gives me time to save for the Loki engines. :p

Here is the link to the certification package: https://drive.google.com/open?id=1i4anunZXsiG7dyIP78nmFN7WJrSK-7uAgVvensHU9ok
 
Last edited:
I have often dreamed of building a full scale Patriot, going to be a plasure to watch your build come to fruition. Question what is the difference between a PAC-3 Patriot and the standard Patriot?
 
Oops. I totally forgot to attach the ORK file. Here it is. You'll notice the fin tabs are quite extended. I'm lining up tabs with bulk plates to form a rigid, inline structural framework. That allows me to only worry about the outer airframe for compressive forces as it hits mach. I want to do something new for me and learn in the process.

You can see the structural support on the 3D unfinished view. Sure, that adds weight, but I plan on testing an adhesion primer on the tube before adding the drain sleeve and epoxy. That will massively cut down on epoxy absorption into the tube which will cut a lot of weight. I just need to test strength. Science! :D

2016-11-29 17_47_39-Rocket (8 inch scratch MEADS PAC-3 L3.ork) 3.jpg

View attachment 8 inch scratch MEADS PAC-3 L3.ork
 
I have often dreamed of building a full scale Patriot, going to be a plasure to watch your build come to fruition. Question what is the difference between a PAC-3 Patriot and the standard Patriot?

PAC-3 is the Patriot Advanced Capability-3 version. A good link for the version comparisons is here.

https://en.wikipedia.org/wiki/MIM-104_Patriot#Variants

The PAC-3 has rear fins that look more like Nike Smoke fins which I like. Also has forward control surfaces kind of like strakes (at least that's what it looks like to me). Several different looks are out. I modeled mine after the Lokheed Martin version shown here:

https://m.lockheedmartin.com/conten...ment-enhancement/mfc-pac-3-mse-photo-01-h.jpg

To save weight, I reduced the forward fin height to more resemble the German AMD variant:

https://meads-amd.com/wp-content/uploads/2012/08/MEADS-Flight-Test.jpg
 
Awesome and Thx. I found this as well for operational range.
PAC-1: 70 km
PAC-2: 96 km-160 km
PAC-3: 20 km against ballistic missile PAC-3 MSE: 35 km against ballistic missile

Wonder if we could try some of their solid fuel??? Wonder if there is a difference..lol

Good luck with your build!
 
That design is really sharp. I like the clean lines of the strikes in combination with those forward positioned square fins. I probably would have never gone there in a million sketches, but it looks great!
 
Data crunching and planning progress has been made.

First off, I wasn't so sure about the strength of the 8" Quikrete tubes, so I did a lot of research on the forces the tubes will see and how they can fail. The end result (as covered in another thread) was the publishing of a tube crush/buckle calculator on Google Sheets.

https://docs.google.com/spreadsheet...FxY3m08tIn-kY2F8sNV7NQd8k/edit#gid=1801452371

Plugging in my numbers, I determined that the tubes alone would buckle under a moderate flight if they weren't reinforced in some way.

In another thread, I've been testing ways to fiberglass the tube without adding crazy amounts of weight. After testing, I've settled on the following method:

1. skip peeling oro sanding away the outer paper wrap. It makes the tube too porous.
2. instead, spray the outside of the tube with adhesion promoter. I used Dupli-color adhesion promoter on a 1' test section.
3. I tested applying 2 layers of polyester drain sleeve with US Composites 3:1 laminating epoxy.

The end result was a pretty robust tube. However, paranoia took over, and I tested again with another section of tubing with 2 x 6oz wraps of fiberglass and one 2oz veil wrap. The weight gain was only a bit higher than the drain sleeve, but it was a bit stiffer.

I both cases, I could not peel off the drain sleeve or the fiberglass in the slightest bit. Using pliers, I could pull it off, but it came with a generous layer of the tube. I'd say I'm happy with how well the epoxy stuck to the adhesion promoter treated tube. Plus, the tube didn't soak up enormous amounts of epoxy; avoiding wasted money and weight.

The final weight for an 8" x 48" Quikrete tube sprayed and fiberglassed was 2.4 kg, or 5.3 lbs.

I used this modified density to recalculate the model.

View attachment 8 inch scratch MEADS PAC-3 L3.ork
 
Next up, I'm figuring out how much harness to use. I'll be using a pilot chute to pull the main from a bag while the nose cone separates with the pilot. I'm estimating this amount of harness. Sound about right?

30' from booster to upper airframe
8' from upper airframe to main chute
6' from nosecone to empty deployment bag
L3 cord diag.jpg

If those are about right, next up is calculating the right chute sizes for the weights.
 
I tried this once before with no luck but here goes again. I started work on a dimensioned drawing fpr the PAC3-MSE some time ago and this is what I have.

The diameter is based on the known diameter of the base PAC3 of 255mm. While the motor diameter increased, the forward control section remained the same. You have to add 6mm to that for the ablative coating to get the true diameter of the finished vehicle. I didn't get to the fin spans but the tip-to-tip distance for the fixed forward fins should be very close to that of the PAC3 since they need to fit into the same space.

The aft control fins fold so that they can fit their extra span into the launch tube.

View attachment pac3mse.topsilk.pdf
 
IMO, way too short for a drogue recovery cord on an 8" bird. Others may disagree, but I think 50-70 feet would be better for an 8" bird.

:2:

Agreed with Justin. For my 8"Gizmo, I used 2-30 foot harnesses. Might've been a 30 and a 25... In either case, at that size, the extra length can only help.
 
IMO, way too short for a drogue recovery cord on an 8" bird. Others may disagree, but I think 50-70 feet would be better for an 8" bird.

:2:

Double ditto. You ejection charges are going to be big to make dang sure that rocket separates and the will dissipate all that energy, too short and all those components are going to be stressed or you can decrease your ejection charge but that's not worth the risk
 
Thanks for the feedback. I bumped up the cord length to 60' for the drogue section. I found a nice deal on 2500# test milspec 5/8" tubular nylon.

https://www.amazon.com/dp/B00PHRUDMG/?tag=skimlinks_replacement-20 for $15 plus $8 shipping. That should fit the bill if I put a sleeve on it to guard from ejection charges.

I have 2 x 30" nomex sleeves already on hand I can use to guard them.
 
Last edited by a moderator:
I tried this once before with no luck but here goes again. I started work on a dimensioned drawing fpr the PAC3-MSE some time ago and this is what I have.

The diameter is based on the known diameter of the base PAC3 of 255mm. While the motor diameter increased, the forward control section remained the same. You have to add 6mm to that for the ablative coating to get the true diameter of the finished vehicle. I didn't get to the fin spans but the tip-to-tip distance for the fixed forward fins should be very close to that of the PAC3 since they need to fit into the same space.

The aft control fins fold so that they can fit their extra span into the launch tube.

Thanks, David for the specs. I tweaked my dimensions a bit to allow for stability when using larger 98mm motors later on. Here are my final dimensions.

L3 fin diag.jpg

Next up... Fin flutter analysis. I haven't built a model (and consequently fins) this big, so I'm checking everything.
 
I'm using Finsim for my divergence/flutter analysis to see if I need to do any tip-to-tip glassing for strength. For fin material, I'm using 1/2" 9-ply Baltic birch. Based on my previous post, here are my Finsim results.

2016-12-09 12_05_49-Fin Geometry For Aeroelastic Analysis.jpg

Since I have a symmetrical fin, I'm assuming the elastic axis and the center of gravity are both the same at the root midpoint (0.5). Someone correct me if I made an incorrect assumption. I only use the following screen to input my CG and EA. The divergence and flutter speeds are more accurately calculated using the Torsion-Flexure analysis.

2016-12-09 12_08_06-AeroFinSim -Fin Aeroelastic Analysis Program By AeroRocket.jpg

The Torsion-Flexure analysis tool says I'm safe from unstable flutter at mach 1.63. U vs g shows max speed for divergence as mach 1.59 (not shown here).

2016-12-09 12_08_31-Torsion-Flexure (2-D) Unsteady Flutter by Theodorsen and U-g methods.jpg

Looking at the hottest 98mm engines I can put in this, the theorhetical best speed I can attain is mach 1.27 using the N5800 from CTI. That's 80% of the safe maximum for divergence, so I don't believe any reinforcement is necessary. For my L3 cert, I'll use a 75mm Loki M1969 sparky which will give a max velocity of mach 0.55. Well within tolerance limits before ripping the fins off.

I'll skip the extra weight and not fiberglass the fins.
 
Last edited:
I forgot to add wild flight check...

Even if I take a 20 degree kick off the rod from winds, the fins won't rip off. They can survive an angle of attack of 20 degrees up to mach 0.3. Well below the speed after weathercocking has leveled off.

2016-12-09 15_00_32-AeroFinSim - Fin Bending Stress Analysis.jpg
 
IMO, way too short for a drogue recovery cord on an 8" bird. Others may disagree, but I think 50-70 feet would be better for an 8" bird.

:2:

I'm one that totally disagrees with having that long of drogue cord. I believe you only need enough drogue charge to reliably open the rocket, it doesn't have to extend to the ends. It only needs to get the drogue chute into the air stream and you will have a good air stream since the rocket will be falling pretty quick.

Don't over do the drogue charge! That causes all kinds of other problems like the main deploying when the payload section hits the end of the cord. My opinion is that you only need a long drogue cord if you have too large of drogue charge or fly drogueless. In either case, you are using a long cord to compensate for a poor recovery design.

The key is the drogue. It has to be big enough to keep the payload section above the fin can. If it doesn't do that, then all bets are off when the main deploys. You might get it to work fine the first 5 or 10 times, but if you deploy the main below the fin can, it will bite you sooner or later.

The other issue with long drogue chords when you are drogueless or have too small of drogue is that the main will open and since the payload is the only section supplying weight, it will almost stop in the air because of the light load when the fin can doesn't hit and foul it. Then after the fin can falls the full length of that too long of drogue cord, it hits the end and puts a huge shock load on the whole system. This is almost never a problem with L1 sized rockets because of material strengths. When you start getting into the larger L2 rockets, it can become more of an issue because the material strengths and larger shock loads give you less margin. At L3, you better design the whole recovery right because doing it wrong can leave you with a fin can falling by itself and the upper section drifting away on an under loaded chute.

BTW, the drogue cord on my 8ft 4" rocket is 15 ft long. It's got 37 flights on it without issue.
/rant off
 
... Then after the fin can falls the full length of that too long of drogue cord, it hits the end and puts a huge shock load on the whole system....

Question for those more experienced...

My model sims a drogue descent rate of 20 m/s (drogue plus airframe drag calculated).

Now, when the main deploys, the booster section will still be falling at an approximately constant 20 m/s regardless of whether the shock cord holding the booster was 10' or 100'? Plus, a longer shock cord would allow more elastic shock absorption. What am I missing here from a physics perspective? I'm not understanding how the booster would accelerate more on a longer shock cord. Then again, I'm possibly overlooking something.

With a properly sized drogue, and the attachment at a point that holds the upper rocket body over the booster, I see the cons of an overly long shock cord as:

1. weight
2. tangling potential

The pros are:

1. more shock absorption at apogee event
2. more shock absorption at main event

Assuming weight isn't an issue on a large rocket where I'm not min/maxing for speed or altitude, I can mitigate tangling with z-folds and tape/bands.

What else to consider?
 
Thinking out loud (is that possible when typing? :tongue: )

The main deploys.
Before snapping open, the upper starts to decelerate.
- short drogue cord tension pulls on the booster a few m/s deceleration before main snaps open
- long drogue cord allows the booster to continue falling at 20 m/s before main snaps open

So the question is, how long does it take for a main to usually snap open? And what's the deceleration rate of the upper body before then? If I can figure out the deceleration rate, the goal would be to size the drogue cord to become taught before the main snaps open while keeping it at a max length for shock absorption (plus more safety margin at apogee event).

Any parachute specs for deceleration before full deployment? And typical times to open fully from a bag assuming normal packing and not loose base jumper packing?

I have the itch to do some maths!
 
Thinking out loud (is that possible when typing? :tongue: )

The main deploys.
Before snapping open, the upper starts to decelerate.
- short drogue cord tension pulls on the booster a few m/s deceleration before main snaps open
- long drogue cord allows the booster to continue falling at 20 m/s before main snaps open

So the question is, how long does it take for a main to usually snap open? And what's the deceleration rate of the upper body before then? If I can figure out the deceleration rate, the goal would be to size the drogue cord to become taught before the main snaps open while keeping it at a max length for shock absorption (plus more safety margin at apogee event).

Any parachute specs for deceleration before full deployment? And typical times to open fully from a bag assuming normal packing and not loose base jumper packing?

I have the itch to do some maths!

Question for those more experienced...

My model sims a drogue descent rate of 20 m/s (drogue plus airframe drag calculated).

Now, when the main deploys, the booster section will still be falling at an approximately constant 20 m/s regardless of whether the shock cord holding the booster was 10' or 100'? Plus, a longer shock cord would allow more elastic shock absorption. What am I missing here from a physics perspective? I'm not understanding how the booster would accelerate more on a longer shock cord. Then again, I'm possibly overlooking something.

With a properly sized drogue, and the attachment at a point that holds the upper rocket body over the booster, I see the cons of an overly long shock cord as:

1. weight
2. tangling potential

The pros are:

1. more shock absorption at apogee event
2. more shock absorption at main event

Assuming weight isn't an issue on a large rocket where I'm not min/maxing for speed or altitude, I can mitigate tangling with z-folds and tape/bands.

What else to consider?

I believe Mark has it right. It's not about shock absorption or acceleration, it's about timing. When you have a long drogue cord and the fin can floats at or above the payload, there is a long time available for the main to deploy, open, inflate and slow or nearly stop the payload section. It's that relative difference in speed that gets you. If your fincan falls at 20 m/s but the main with only the payload section loading it drops at 4 m/s that's a pretty significant shock to the system when the fin can hits the end at a relative 16 m/s. With a shorter cord, the main may not be fully open or slowed the payload much and the difference when the fin can hits the end of the cord might only be 5 - 10 m/s. Of course the proper drogue that keeps the fin can below the payload will pretty much prevent that type of shock loading.

That is also why I said the problem isn't usually apparent in L1 rockets. Their fin cans are lighter, the shock cords and attachment points are relatively stronger and the safety margins are a bigger. When that fin can goes from 3 lbs on a L1 to 30 lbs on a L3, your shock loads get much larger very quickly.

If you want to do math, calculate how much force various weight fin can impart at various speeds. You can calculate the drop weight of the main chute with just the weight of the payload section and use that as the difference in the fin can drop rate to calculate the impact speed.
 
Thanks. TAP is triggering the project now. We'll see what changes will need to be made.

In the meantime, I'm turning all those 8" wood circles down by an 1/8th inch so they will fit in the tubes as centering rings and bulkheads . 10 down. 10 to go.
 
If you want to do math, calculate how much force various weight fin can impart at various speeds. You can calculate the drop weight of the main chute with just the weight of the payload section and use that as the difference in the fin can drop rate to calculate the impact speed.

I think so. If the lower half is dropping at the calculated speed of 20 m/s and snaps to a slower 7 m/s, the force will be about 41 lbs force. Still it too bad.
 
I got all my centering rings sanded to fit in the Quikrete tube, and I use a 98 mm hole saw to cut the hole for the motor mount. I had to sand the hole a tiny bit larger to get a tight fit on to the LOC motor mount tube.

20161221_192650.jpg

The centering rings will all be glassed with a layer of drain sleeve and then doubled with a second ring to create a core of two layers of fiberglass. I did a test with a couple circles of this 1/8th inch burch plywood, and it came out like steel. The centering ring sandwich will be:

fiberglass
wood
fiberglass
fiberglass
wood
fiberglass

Here's a picture of the test fit with the centering rings and the axial strengthening strakes. The fins will go in the lower large gap. Shorter strakes (not shown) have been cut to fit between the thrust plate rings and the lower centering rings. A coupler will butt up against the topmost rings to create a continuous reinforcement from thrust plate up to the payload bay.

20161221_200419.jpg

Next, I need to cut the fins from the 1/2 inch birch plywood and start the compound bevel. I have a template and guides for the primary and secondary bevels, so hopefully they'll all come out really close to the same shape. After sanding, those will get a layer of glass for plenty of flutter safety margin.

I don't want to get too far ahead until my TAP has a chance to provide feedback.
 
Yay. I got the green light to proceed from my first TAP who's local. Need to find a good second TAP to help advise now. The hunt begins... I reached out to all the local(ish) TAPs to see who might have the time to advise.
 
Last edited:
I had time tonight to create the epoxied centering rings. The starting material was 3-ply 1/8th inch birch plywood. I sourced the 8" circles from https://www.craftparts.com/wooden-circles-sign-p-4028.html at $13 for a bag of 10. Since they were 8", they matched the outer diameter and had to be sanded down. I did this by marking the center, drilling a small hole, and putting a bolt with large washers and nuts through to clamp them all together. The bolt was put in a drill and run across a clamped sanding block until they fit the ID of the Quikrete tubes. Worked surprisingly well.

The inner hole for the 98 mm motor mount was cut out using a 98 mm hole saw. https://www.amazon.com/gp/product/B000E8FP6M/?tag=skimlinks_replacement-20

I marked the grain direction on each ring since they will be layered in perpendicular directions when making the fiberglassed sandwich.

20161228_205825.jpg

The rings were inserted into 8" diameter polyester drain sleeve for the reinforcing fabric; making sure the grain direction registration marks were perpendicular. The rings are in pairs since they will be folded over as a sandwich after the epoxy is added.

20161228_211715.jpg

I used US Composites 3:1 medium laminating epoxy. The polyester drain sleeve turns moderately transparent when wetted out. Excess epoxy is scraped out using a teflon scraper since it only adds weight and not strength. Unfortunately, I didn't deglove and take an in-progress pic. All the epoxy work was done on top of a sheet of mylar to keep from sticking to the board that I've sanded smooth as a good base for epoxy work.

The folded over sandwich was covered on top with mylar to and weighted down with a board for compression.

20161228_220905.jpg

Tomorrow, I'll rip off the mylar and trim the edges with a razor.
 
Last edited by a moderator:
I think so. If the lower half is dropping at the calculated speed of 20 m/s and snaps to a slower 7 m/s, the force will be about 41 lbs force. Still it too bad.

I'm not sure I have all the figures right, but assuming a ~29 lbs, or 13kg fincan and it goes from 20 m/s to 6 m/s or 14m/s to zero in 0.1 seconds when hitting the end of the shock cord. The acceleration would be 14/0.1squared) or 1400, your formula would be F = 13 * (14/(0.1 *0.1)) = 13* (14/ 0.01) = 13 * 1400 = 18,200kg or 40124.132 lbs of force. It would be much more then 41 lbs.

That's why I say you almost never see the failures in L1 or smaller L2. Assuming a 5 lbs fincan or 2.27kg, the force would be 2.27 * 1400 or 3178kg or 7006.291 lbs, which is much easier to handle. a 2.5 lb. fincan would only impart about 3500 lbs of force.
 
Last edited:
Back
Top