I've been wondering how a rocket fuselage can support all the weight of the upper stages when it is only made of a cylinder of very thin sheet metal. (Especially considering acceleration, vibration and aerodynamic force.) A few rockets have relied on internal pressure for strength, but these are not the ones I am talking about. In general I would appreciate any insights on the engineering principals used.

My specific question is: Why aren't rockets built with truss structures inside their fuel & oxidizer tanks to increase structural strength?

  • 1
    $\begingroup$ The message I'm getting is that a simple cylinder is the most efficient approach. I assume it would loose strength rapidly if it is deformed. Deformation may be more of an issue in the age of reusable rockets. Deformation may be cause during return by unequal heating, or off axis arial maneuvers. The Starship will reinter on it's side, experiencing lateral G forces, uneven heating, aerodynamic buffeting as it transitions to engine first orientation, plus high stress points at it's 4 control surfaces. So in all these cases I might expect some internal structure to keep the cylinder a cylinder. $\endgroup$ Aug 15, 2019 at 4:33
  • 1
    $\begingroup$ To clarify, I realize rockets built as cylinders are working, so the idea was to create improvements. Also I realize mass ratios are the name of the game, I was not suggesting adding weight, rather a trade off of being able to make the skin lighter if there was internal structure. The other big cylindrical vehicles people have made are dirigibles. They are full of trusses both linear and lateral as circular trussed ribs. They are used to maintain the shape. $\endgroup$ Aug 15, 2019 at 4:37

6 Answers 6


There's almost nothing to be gained by a truss. The load being applied is along the axis of the tank. A simple hoop of material is very strong in this orientation. (Try it with a piece of paper, you'll be surprised at how much it can hold--just keep the weight even!) A truss in the tank would only help against loads off axis--and you don't want those in the first place!

  • 17
    $\begingroup$ An even better comparison would be to a soda/beer can. Easy to crush sideways, but you have to really stomp on them to flatten them lengthwise. $\endgroup$
    – jamesqf
    Aug 13, 2019 at 17:30
  • 2
    $\begingroup$ Even harder to tear a soda can up lengthwise without nicking it :) $\endgroup$ Aug 14, 2019 at 8:33
  • 10
    $\begingroup$ " A truss in the tank would only help against loads off axis--and you don't want those in the first place" ...unless you're trying to do something like air launch, that is. Which, AIUI, is a non-trivial issue for folks trying to do that, as the extra structural reinforcement needed to let you hang your rocket from an airplane eats into the meager mass savings you get from hauling it above the lower atmosphere in the first place. $\endgroup$ Aug 14, 2019 at 10:56
  • 11
    $\begingroup$ Exactly. These types of questions should always start by asking what problem this 'solution' is trying to solve. What is wrong with the rocket that we are considering adding mass to the structure? The perfect rocket would be 100% fuel with a payload on top. Sadly, we need to add some non-fuel mass to contain the fuel, react it, and direct the exhaust. The rocket equation is punishing - any mass you can delete from the design is a major, major win. If you have a working rocket the question shouldn't be what mass can we add, it should be what mass can we remove and have it still work. $\endgroup$
    – J...
    Aug 14, 2019 at 14:15
  • 8
    $\begingroup$ Going further, it is enlightening to consider that the design of modern rockets has emerged from exactly this type of process - rockets have been designed, tested, evaluated, and mass is continually being deleted at any possible opportunity. Rockets look like they do because engineers have ruthlessly shaved away any mass that doesn't absolutely need to be there. This is a big field where less is more. $\endgroup$
    – J...
    Aug 14, 2019 at 14:18

Because they don't need to be.

Clearly the current design of rockets can be successful. So adding truss structures to the current design would add weight for no reason and take away from the payload capacity.

  • 56
    $\begingroup$ Although I agree with your answer, I'm afraid that there's a bit of a circular reasoning in it. IMHO the question is why the current design is the best one, and "because it is so" doesn't really answer it. $\endgroup$
    – TooTea
    Aug 13, 2019 at 11:26
  • 3
    $\begingroup$ @TooTea I try to answer the question that is actually asked, not what I think the question ought to be. Note the word 'increase' in the question, implying a change to current designs. $\endgroup$ Aug 13, 2019 at 11:38
  • 10
    $\begingroup$ @TooTea It's not circular at all. The question is, why don't we add trusses to the tanks to increase their strength, and the answer is because they're already strong enough without trusses. $\endgroup$ Aug 13, 2019 at 17:38
  • 10
    $\begingroup$ In defense of @TooTea saying "additional engineering is unnecessary because the current engineering works (most of the time)" is not really answering the question. Just because a certain engineering solution is "good enough" for most purposes or for some level of risk, doesn't mean that better engineering solutions are not possible. The original question is asking why a hypothetical change to the solution wouldn't make a hypothetically better rocket. Answering that the current solution is "good enough" is not really illuminating and does not really reflect the drive of human ingenuity. $\endgroup$
    – Daniel
    Aug 13, 2019 at 19:23
  • 9
    $\begingroup$ -1 I don't understand how this is an answer. Adding weight with a truss might allow you to save weight elsewhere - overall this could (in theory) result in a reduction of weight, which would be a big benefit. This answer fails to explain why that isn't accurate. Claiming that if something was useful we'd have done it by now is a claim invalidated by just about every innovation in history. $\endgroup$ Aug 14, 2019 at 14:09

Most modern rockets do rely to some extent on tank pressure for strength. The tankage needs to be pressurized in any case to drive the turbopumps without risk of cavitation, so the structural strength benefits come for free or nearly so.

I'm not certain what you're envisioning when you say "truss structures". There are usually strengthening ribs along the interior walls of the propellant tanks -- welded in in rockets like Zenit and Falcon 9, milled "isogrids" in Atlas V and Delta IV. This provides enough strength to handle the g-loads encountered in the ascent (often as high as 6-g depending on the launcher and mission details), so there's no need for any cross-tank support structures.

  • 1
    $\begingroup$ I remember some rockets have mini-tanks inside with inert gases to compensate the pressure loss due to fuel getting used up. Is this also done to support the structure (at least inside the atmosphere) or "just" to aid with fuel flow? $\endgroup$
    – DarkDust
    Aug 13, 2019 at 9:39
  • 4
    $\begingroup$ @DarkDust Most rockets do, unless they use a fancy approach called autogenous pressurization where they feed some heated gaseous propellants back from the engine into the tanks. Keeping the tank pressure reasonably constant is crucial to prevent pumps from cavitating. $\endgroup$
    – TooTea
    Aug 13, 2019 at 15:03
  • $\begingroup$ @DarkDust One thing to keep in mind is how rockets are stacked - the mass of the other stages is very small compared to the mass of the currently active stage, until you're close to empty on fuel. Since you need to actually keep relatively constant pressure in the tank to have the fuel pumps work, the strength doesn't change much as the tank empties out, and the strength of the tank is designed to hold up the entire weight of the upper stages at the acceleration achieved by the now-lightened rocket - if the strength isn't enough, you just make the walls of the tank thicker. $\endgroup$
    – Luaan
    Aug 14, 2019 at 10:32

Because it would be an inefficient way to handle the loads.

Let's say your rocket is a simple cylinder with engines at the bottom (no strap-on boosters or fins that might actually need extra structure to attach to and transfer the loads). Such a rocket will be subject to two main kinds of loads:

  • axial compression (engine thrust vs dynamic pressure of ramming into air head-on)
  • bending/shear by aerodynamic forces (flying at nonzero angle of attack causes the body to generate some lift)

As hinted in other answers, compression is easy to handle with what you already have: the skin of the cylinder. You just need something with a sufficient cross-section that won't buckle easily, and a big metal pipe is a good match for that requirement. And guess what, you already need that pipe to keep your propellants in.

The bending is a bit more tricky (and it also comprises vibrations of various frequencies), but a truss won't help very much with that. For a truss to resist bending, you need to make it wide. A single rod on the axis of the rocket won't help. And as you make it wider and wider, it will become stiffer against bending, until it finally is as wide as the entire cylinder. That means you've found the optimal arrangement: strengthening the walls.

  • 4
    $\begingroup$ You can look an nature to see how you properly design against bending. The reference design is bamboo; the optimal strengthening is not a truss but a disk. $\endgroup$
    – MSalters
    Aug 13, 2019 at 14:29
  • $\begingroup$ Or a femur. This truss structure could be inside the skin, making it thicker but not much heavier, honeycomb sandwich like $\endgroup$
    – user721108
    Aug 15, 2019 at 8:52

I worked on the Atlas rockets which had pressurized tanks (balloon tanks) for structural rigidity/strength. For transport, the Atlas went into a truss to hold the rocket in tension to keep it from collapsing. On the pad, it needed to be pressurized. I once saw a pic of a retired Atlas on display, but the air pump had failed and it crumpled like a soda can. (I searched the internet but can't find it now.)

Also, we were warned not to touch the stainless steel skin with our bare hands, because the skin oils could cause a weak spot and mission failure. Not sure if that was true, but when you consider it gets filled with cryogenics, who knows.

  • 2
    $\begingroup$ Not touch the inside or outside of it? Any large object standing outside a building is potentially subjected to more nasty random contamination than skin oils.... $\endgroup$ Aug 14, 2019 at 8:38
  • 1
    $\begingroup$ This doesn't answer the question. $\endgroup$ Aug 14, 2019 at 12:58
  • 2
    $\begingroup$ New question prompted by this answer Will some rockets really collapse under their own weight? $\endgroup$ Aug 15, 2019 at 16:36
  • 1
    $\begingroup$ @rackandboneman A special coating fluid was sprayed on the outside of the Atlas to protect it by displacing water away from the skin. This Water Displacement formula took 40 attempts to develop: WD-40. (The story sounds too good to be true, but I haven't seen it debunked, and Snopes confirms it.) $\endgroup$ Aug 15, 2019 at 17:21
  • $\begingroup$ Was thinking you'd need "bird crap displacer 40" to protect a corrodable object outside :) BTW, I would normally only expect a martensitic stainless, or a semi-stainless steel to care about normal touch even in the slightest? Given many households pressure cook thinks full of salt and oil at 121°C and 15 psi overpressure in austenitic stainless vessels routinely :) $\endgroup$ Aug 15, 2019 at 18:57

A problem with trusses inside a liquid is that you have the possibility to set up currents and cavitations in the fluid that can cause it to move in ways you don't want. Without the trusses, the fluid moves (moderately) smoothly in the direction of the nozzle. With the trusses, it can set up eddy currents and (in extreme cases) voids in the liquid.

When you have a pressurized vessel, you don't really want any extra pressures that you can't expect. The collapse of a void/cavitation can set up pressure waves that may cause over-pressure in a weak spot in the hull or, in this case, uneven burning of the fuel.

In an extreme case of many trusses, you might end up with an unintended baffle system, where it restricts the fuel flow so much that it slows the burn, potentially causing the engine to stall. Semi-tankers use baffles to help prevent sloshing so they can brake and accelerate easier. In the case of a rocket, you aren't stopping and starting a lot, so sloshing shouldn't be a problem, especially with essentially a single force acting on it in the direction the fluid needs to go.

Also, rockets have a problem named after them: The Tyranny of the Rocket Equation. It says that the more weight you have, the more fuel you need and the more fuel you need, the more fuel you need to lift the more fuel you need to lift the weight of the fuel you need to lift the other parts of your rocket. Simple, right? Trusses add weight to the rocket, so if you can do without them, you're better off. As other people mention, the skin of the tank and ship are more than enough to deal with the pressures and forces involved. If they weren't, the engineers would have added the necessary trusses. ;-) Or used a stronger material.

When you have a reusable tank that is recovered after a mission, you have to inspect it inside and out. With more surfaces and joints/seams/welds to inspect, you increase the time, effort, and cost to do the inspection. If there are any coatings to prevent the fluids from eating away the tank, you increase the cost of originally applying and any need for reapplying them by adding structure to the inside of the tank.

  • 2
    $\begingroup$ Many tanks contain baffles to reduce slosh. A truss would help rather than hinder in this regard. Most baffles however are rather lightweight and offer little structural stiffness. $\endgroup$ Aug 14, 2019 at 17:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.