1/26th Scale Space Shuttle (full stack)

The Rocketry Forum

Help Support The Rocketry Forum:

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

sambatterman

Well-Known Member
TRF Supporter
Joined
Feb 9, 2022
Messages
294
Reaction score
502
I'm starting a new long-term project. I want to build a 1/26th scale Space Shuttle full stack. Meaning an external fuel tank (ET), two separating solid rocket boosters (SRBs) and a gliding home orbiter. This will be ambitious and will probably take a number of years to complete with varying degrees of success along the way.

Why the Space Shuttle?
If you look at what's happening in our hobby, there is a new catalyst that's starting to shake things up: flight computers and software. Now, I get that altimeters and timers have been around for a long time, but I'm not talking about that. I'm talking about flight computers that are beginning to add capabilities that we thought were unthinkable not too long ago. Most of this has come from Joe Barnard's amazing mind (bps.space), but others are also working on this - first in low power rocketry (as Joe did) and now it's expanding into mid to high power rocketry. BPS Space is starting to get to a place with hardware, software and material science where thrust vectored flight is coming - this will eliminate the need for fins and will add a number of new capabilities to our hobby: propulsive landing, better staging, etc.

This will lead to new vendors, new tools, new materials (like ceramic thrust vanes) and new capabilities that only software can bring (roll programs, fly by wire, etc.).

One of the things I love about rocketry is it's both wide and deep - you can sit at low power rocketry and develop thrust vectored flight or go to level 3 and get to space. It's as big as you want it to be, with no pressure to go either way.

The Space Shuttle has been a challenge in our hobby ever since Estes created the first one in 1976 with cruddy, vacuum formed plastic parts and an offset motor engine in the external tank. The model was complicated and unwieldy, but decidedly cool. While it was expensive, it was cool to spend $25 and have the most advanced model rocket available. There has been a lot of work in this area by people like Andy Woerner and George Gassaway - both people I have studied for the last two months trying to talk myself into this project. We all stand on the shoulders of giants and Andy and George (and many others) have paved the way for this.

The Space Shuttle has all the challenges that the hobby/sport has been chasing for a long time:

  • Separation of components into flight - requiring a timer and other things like passive and active traction on the boosters.
  • A complex center of gravity - the orbiter, ET and SRBs change a simple measurement into something devilish.
  • The need for flight controls acting during boost, separation and recovery - perhaps a little beyond what a human can deal with.
  • Flight controllers that have been fire-tested by other hobbies. A great example of this is Ardupilot - an auto pilot for drones, fixed wing powered and unpowered flight and helicopters and boats and submersibles (https://ardupilot.org/).
1/26th scale is big enough to work on the things that really complicate this thing without throwing it all out from a weight perspective (a serious consideration). I will share more on the dimensions soon, but the components needed to pull off this scale readily exist and are relatively cheap - at least for the ET and SRBs. The orbiter will be expensive to build - not just because of the form or the vehicle, but the function of elevons, rudders and propulsion. Trying to land something like the shuttle safely with little or no human intervention represents a "moonshot." I expect there to be a number of sensors (lidars, GPS units, altimeters, etc.) with an Arduino computer driving servos controlling split elevons, the rudder and the body flap directly under the motors.

There's a lot to prove to RSOs and I expect there to be a lot of feedback on this build. That's welcomed as long as it's constructive.

In the meantime, check out my 1/2 scale ALCM Boeing Cruise Missile that's getting close to being finished in the next month or so.

See you on the journey...
 
Last edited:
I will be watching with great interest!

I've thought of the same for many years now and had the Sheri's Hot Rockets Shuttle kit from Red Arrow (which I REALLY wanted to build), but sold it after losing my job in 2020.

For a while I was also talking to the owner of Hangar 11 about the larger shuttle stack he was developing and by the time I tried to buy it, he stopped selling rocket items and no longer returned my correspondence.

I commend you sir for not being one to shy away from challenges! :) đź‘Ť
 
So why exactly 1/26 scale? I personally would try for 1/100 to match with the Saturn kits, but 1/24 would match with common plastic model kits.
 
Why 1/26th Scale?
As I studied through the drawings and dimensions for this thing, I decided that I wanted to go with high power for this project. Level 2, which is where I am in certification, gets me access to the lift I need, while still affording a larger platform. Larger RC models are easier to control and are less buffeted by wind, etc. In the case of the orbiter, it is self-described as "flying like a brick." So, larger surfaces on the delta wing and larger controls will be a benefit, I think.

The External Tank (ET)
At roughly 1/26th scale, the External Tank (ET) is a 12" Sonotube at Lowes/Home Depot with an 8" interior tube for recovery. The ET would be roughly 71" long with a nose cone that is around 20 1/2" --- at $35 to buy both these tubes it's the cheapest way to get started with most bulky part of the stack. Yes, I will need to glass the Sonotube and there will be internal support with plywood center rings, etc. Not to mention a number of 3D printed parts - especially the nose cone. The attach points from the ET to the orbiter will be figured out with 3D printing, but once that is understood will likely be fabricated in aluminum. I'll be working with a local machine shop on that as I do not have that expertise.

The Solid Rocket Boosters (SRB)
The SRBs, of which there are two, can be approximated with 5.5" Blue Tubes - in all drawings, the ET is basically the diameter of two SRBs, so this works out quite well. While Blue Tube isn't the cheapest thing around, I have used it in my 1/5 scale Nike Hercules and 1/2 scale ALCM Cruise Missile. It is easy to work with and finish and is strong to start with. At $105 each, the SRBs will be 72" long - this includes the nosecone (3D printed), skirt and motor retention. These too will be glassed for strength. The connection between the ET and SRBs will be of interest to many - how to provide passive and active retention, while still allowing separation before apogee. I'm studying some work done by John Coker on his Ariane 6 for using solenoids to separate the boosters.

The Shuttle Orbiter
The hard part. The orbiter, a 1/26th scale, is around 55" long, with a wingspan of exactly 36" - this means the fuselage is 8 1/2" wide and the rudder stands 12 1/2" tall off the fuselage. Each wing extends 14" away from the fuselage. While the model itself will be birch and balsa, with carbon fiber reinforcements, I'm wondering if I should take the time and make it a plug and go for the lightweight fiberglass shell. This makes it easier to replace in the very real possibility that I will lose orbiters to reach the stated goal of the project.

Other Considerations
Getting to the field to fly the thing will probably require a rented mini-van or truck.

Propulsion needs to be adequate but not so expensive that this is prohibitive to fly. Right now, I'm thinking the SRBs are L2200s and the orbiter has 3 "J's" - I'm still figuring that out. This is a low and slow flight - I'd be happy with apogee in the 2500-3000' range. I'll be sharing some OpenRocket sims soon. While I can sim each component, the entire stack is not something that OpenRocket was designed for.

Avionics will involve a lot of components:
  • Timers and Altimeters for separation and recovery.
  • Arduino for main processing and interface with sensors.
  • 2 solenoids for SRB separation.
  • 2 solenoids for Orbiter separation.
  • The flight computer will be based on the Ardupilot platform.
  • Barometers
  • Airspeed sensor
  • IMU (gyro)
  • GPS sensor - probably a few to determine orientation
  • Rangefinders - for determining distance to the ground, flare, etc.
  • and, of course, servos for articulating the control surface - split elevons and split rudder and electric landing gear.
 
Last edited:
Maybe I'm wrong, but I think your problem is going to be weight. Even if we assume you've got a fairly large motor in each SRB, you're also carrying an extra "big tube" that is dressed as the external tank, and then you're carrying the shuttle model, WITH all it's associated electronics for flight/glide post-apogee PLUS whatever electronics are needed to make the whole thing stable during launch without additional fins. If you're gimbaling the thrust for active control during the "up" portion of the flight, making that strong and yet light enough to also carry the orbiter and the tank, that's the challenge right there.
 
Maybe I'm wrong, but I think your problem is going to be weight. Even if we assume you've got a fairly large motor in each SRB, you're also carrying an extra "big tube" that is dressed as the external tank, and then you're carrying the shuttle model, WITH all it's associated electronics for flight/glide post-apogee PLUS whatever electronics are needed to make the whole thing stable during launch without additional fins. If you're gimbaling the thrust for active control during the "up" portion of the flight, making that strong and yet light enough to also carry the orbiter and the tank, that's the challenge right there.
Weight will definitely be something to keep under scrutiny. With L2200s and 5:1 safety, the entire stack needs to be less than 95 pounds. This is not including the lift from the orbiter. I'm still working through the lift - it's complicated. During this project I guarantee you will hear a common phrase from me - "I don't know exactly. I'm working on it." Additionally, by using 75mm MMTs on the SRBs and 38mm MMTs on the orbiter, I'm trying to create options as I work through the right combination of motors to get the thrust needed without tearing the thing apart. One last point, I'm trying to avoid sticking a big motor in the bottom of the External Tank, but if that has to happen for safety reasons, I will.
 
Last edited:
If you can fit 54mm MMTs on the orbiter, J180s have a nice flame color that is a lot like LH2 flame out of the SSMEs.
 
I decided that I wanted to go with high power for this project. Level 2, which is where I am in certification, (...)

Other Considerations
Propulsion needs to be adequate but not so expensive that this is prohibitive to fly. Right now, I'm thinking the SRBs are L2200s and the orbiter has 3 "J's" - I'm still figuring that out.
Someone can correct me if I'm wrong, but I believe you are going to need a Level 3. That cluster of motors is easily going to push you into the "M" range.
 
Someone can correct me if I'm wrong, but I believe you are going to need a Level 3. That cluster of motors is easily going to push you into the "M" range.
You're probably right - I will probably use my Nike Hercules to get that done after an initial shake down on an L2200. This is multi-year project so I'm thinking I have the time to get there.
 
You're probably right - I will probably use my Nike Hercules to get that done after an initial shake down on an L2200. This is multi-year project so I'm thinking I have the time to get there.
Cool, you should be fine once you get that squared away.

Just noticed from your sig that you're also an MDRA member. Looking forward to seeing the progress on this and eventually seeing it fly.
 
If you can fit 54mm MMTs on the orbiter, J180s have a nice flame color that is a lot like LH2 flame out of the SSMEs.
Flame color is a nice benefit, but likely will have 54mm MMTs on the orbiter to add additional thrust. I can always use adapters to go down, but I can't go up. :)
 
OpenRocket - Modelling the Stack

OpenRocket is a fantastic program. I rely on it for lots of things: estimated weight, CP/CG calculation and visualization and estimated altitude. I've never made anything complex enough to challenge the software.

This project seems to be there - complex.

When I took the Space Shuttle stack and modelled it, I got the initial weight - around 70 pounds without motors. This is modelling the materials correctly (fiberglass, plywood center rings/bulkheads, etc.) and including the recovery items (parachutes and shock cords). I even added 5 pounds for the avionics hardware (computers and sensors). So, I think this is a pretty good look at the initial weight of the stack. Will it go up? Undoubtedly. But my experience is that this is usually pretty close.

The currently available OpenRocket program does not handle the "stacking" well at all. So, when the various components of the stack are described, they are laid out linearly in the design surface with no real ability to change the orientation. This had me perplexed for a little while, but I ran into another rocketeer on Rocketry Forum named Jeff Lassahn who is working on a 1/150th shuttle stack. He had some great recommendations to help me get this stack into a more realistic formation.

I want to share that here, so that others can model complicated stacks as well.

linear stack.png

First, you need to download the beta (at least at this point in the lifecycle). You'll need version 22.02 beta 05 or greater. This version includes a new building component called a Pod Set that allows you to relate and orient components one to another. The trick here is that the pods are linked to a "dummy" body tube (identified as #*ref). These tubes are zero length and height. Then the components are related to the pod and allows for changes to orientation to the External Tank.

Here's what the design explorer looks like to give you a better idea.

explorer.jpg

So now after using the pods, we can visualize the stack like this. Before anyone rips into me, yes, the motors on the orbiter will be canted upward, you can't do that in the software that I could see.

Stack 1 65 pnds.jpg

And here is a top-down view.

top down stack 65 pounds.jpg

There are a number of things that came out of the stacking exercise that I want to talk about, but that will have to wait for another thread.

Here's the ORK...
 

Attachments

  • SpaceShuttleStack.ork
    3.3 KB · Views: 0
Last edited:
Can you post that ORK file please?

Actually, I think I can comment without even seeing it, based on what you've posted above: right now you are treating everything as a separate stage, which is not optimal. It might work, but you're not using the tool in the most efficient way.

Instead, the two side boosters should use the "Boosters" component:
1670518466938.png
These can be attached to a body tube, which in this case would presumably be the external tank. They are much the same as pods but can be modeled to "fall away", which is what I imagine yours will be doing.
1670518545048.png
I could imagine the entire stack would be modeled as two stages: the orbiter, and the external tank + boosters.

If you do it this way you will have a simpler design in OR with fewer phantom body tubes. Ultimately you will probably need to do that trick only once, to place the orbiter on top of the booster stage.

If this is not clear I can demonstrate it in more detail when I have the file.

This is certainly a great way to exercise the new OR version to its fullest!
 
Last edited:
Can you post that ORK file please?

Actually, I think I can comment without even seeing it, based on what you've posted above: right now you are treating everything as a separate stage, which is not optimal. It might work, but you're not using the tool in the most efficient way.

Instead, the two side boosters should use the "Boosters" component:
View attachment 549885
These can be attached to a body tube, which in this case would presumably be the external tank. They are much the same as pods but can be modeled to "fall away", which is what I imagine yours will be doing.
View attachment 549886
I could imagine the entire stack would be modeled as two stages: the orbiter, and the external tank + boosters.

If you do it this way you will have a simpler design in OR with fewer phantom body tubes. Ultimately you will probably need to do that trick only once, to place the orbiter on top of the booster stage.

If this is not clear I can demonstrate it in more detail when I have the file.

This is certainly a great way to exercise the new OR version to its fullest!
Thanks Neil! I will give that a shot. Appreciate your feedback! I presume this will resolve the thrust calculation issues I was having with modelling everything as a separate stage.
 
I understand wanting to have motors in the shuttle itself, but putting clusters in a gliding model is a LOT of tail weight and corresponding nose weight and will make your wing loading/glide a lot worst, you might want to try having the orbiter unpowered at first and see how it flies, if you keep weight light for the wing area they can glide quite nicely actually, this coming from someone who has built many many R/C orbiters with hundreds of flights on them. It will also reduce offset thrust and make your guidance problem less complex.

Frank
 
Using OpenRocket Boosters for the Win!

Wow, what a difference! Neil_w's advice to use the booster component instead of pod sets for the SRBs makes this a lot easier to maintain.

1670520577995.png

And results in a nice visual...side and back views...

1670520624232.png1670520852500.png
 
I understand wanting to have motors in the shuttle itself, but putting clusters in a gliding model is a LOT of tail weight and corresponding nose weight and will make your wing loading/glide a lot worst, you might want to try having the orbiter unpowered at first and see how it flies, if you keep weight light for the wing area they can glide quite nicely actually, this coming from someone who has built many many R/C orbiters with hundreds of flights on them. It will also reduce offset thrust and make your guidance problem less complex.

Frank

Thank you, Frank - I was hoping you were watching this thread. I totally agree and get your point. I'm working with Neil to get a better read on the thrust sim for the stack. The SRBs have L2200s, which should be able to lift the stack. The question of using the control surfaces of the orbiter, especially at boost with or without motors on the orbiter is another conversation.
 
Thank you, Frank - I was hoping you were watching this thread. I totally agree and get your point. I'm working with Neil to get a better read on the thrust sim for the stack. The SRBs have L2200s, which should be able to lift the stack. The question of using the control surfaces of the orbiter, especially at boost with or without motors on the orbiter is another conversation.
Of course I only have experience with the orbiters as standalone models and the largest I've done is around 140th scale but it was only 15 oz so very light it was around 36 in Long and a 25 inch wingspan, I prefer elevon controls but I know George gasaway likes rudder elevator better and from his experience with the full stack when the orbiter is mounted the rudder had reversal and had the opposite effect compared to when it was in glide mode I'm sure he'll chime in here with links to previous information and postings.
 

Attachments

  • WP_001102 (1).jpg
    WP_001102 (1).jpg
    150.6 KB · Views: 0
  • WP_001107.jpg
    WP_001107.jpg
    144.3 KB · Views: 0
  • 20150720131210 (7).jpg
    20150720131210 (7).jpg
    34.1 KB · Views: 0
One Wild and Crazy CP/CG and Verne Estes' "Force Ghost"

It's becoming clear that a huge challenge of this will be how to manage the thrust, weight and CG/CP of this stack.
First, in the OpenRocket model, the Orbiter stage and ET stage need to be in the proper order (thanks Neil). Now that I got the stack arranged correctly, trying to make sure this thing has enough thrust and getting the CG/CP right is an early priority.

Without any intervention, this thing will be sky writing and probably banned from the flying field. (Thanks, Neil, for the term "sky-writing" - I'm going to use that from now on to discuss bad CP/CG relationships :) )

Weight in the orbiter from boosters is another concern, which Frank brought up in the thread above.

As I started working on this thing, it's clear to me now why Verne Estes put those ugly wings on the SRBs in the 1976 Estes Space Shuttle model. Even as a young kid it looked like a huge "cheat" to me. But without them, the CP can never be dealt with.

Without putting any nose weight or "faux" wings on the thing, the CP and CG feel completely out of place.
In an attempt to solve "everything" I removed the Orbiter's motors completely. Relying completely on the SRBs for thrust.
The CP is at least 1.5 calibers ahead of CG - basically it's a bomb.
1670524539653.png

So, by throwing some small fins on the SRBs we get the CP to where we would expect it to be for any other model.
I feel like the "force-ghost" of Verne Estes is sitting next to me on this...
1670524399224.png1670524443934.png

I'd prefer not to do the fins, but It's interesting what they do to a finless model and how they so directly impact the CP.

So, I decided to play with weight in the nose to influence the CG. I put 5 pounds in the nose of the orbiter - and I expect there to be weight there no matter what, but 5 pounds might be too heavy. And twenty pounds in the ET nose, which would not be unusual for a model of that size and thrust. This moves the CG in the right direction and in the image below the CG to CP is 13.3 inches apart (remember the ET diameter is 12"). This brings the weight of the stack to roughly 90 lbs. 110 lbs. with motors.

This is making me wonder whether putting a motor in the ET helps with thrust and keeps motors out of the orbiter.

1670526114116.png
 
Last edited:
To be fair, I seriously doubt that OpenRocket can account for active stabilization - i.e., thrust vectoring and/or gimbaling, or is that portion of this project being abandoned? Modern "real" rockets do not use fins of any type, because somewhere down the line, engineers figured the drag wasn't worth the fuel costs, it was easier to gimbal. How stable would OpenRocket think the real Falcon9 is?
 
Or, thought another way: If you left the bottom of the External Tank "open", would that create enough base drag to let you get away with no fins?
 
Or, thought another way: If you left the bottom of the External Tank "open", would that create enough base drag to let you get away with no fins?

I think that gimbaling the SRBs is literally the only way to do this without adding fins of any type. The real question is can this gimbaling be static or must it be dynamic. In other words, vectoring the thrust in a "hard coded" way, or using software to move the nozzles or vanes to react to what's happening - clearly sensors and software (bps.space stuff). My reflex is it's dynamic.
 
Sticking a Motor in the External Tank?

Just a quick sketch to see what happens to the stack if I put a K-1000 in the ET. It doesn't impact apogee in any degree.
My Nike Hercules is roughly the same weight and with a single L2200 the predicted altitude is 3000 ft. Just to check, I put an N2000 into my Nike Hercules Sim and it went from 3500 ft on an L2200 to 8500 ft on an N. So, I'm wondering if OpenRocket doesn't know exactly how to deal with something like this.

1670528589484.png

Trying to make my point - if I stick an N2000 in the ET instead of a K-1000, I get another 100' in altitude.

1670529176156.png
 
Last edited:
Check your flight side profile graph. Since the Cg and Cp aren't in line, I'd bet it's simulating it doing a big old loop and plowing straight into the ground.
 
Back
Top