Community > Forum > Official Interorbital Systems Forum > Modular Rocket Design

Modular Rocket Design

Posted by: TheFlyingkiwi - Sat Aug 01, 2009 10:59 am
Post new topic Reply to topic
 [ 10 posts ] 
Modular Rocket Design 
Author Message
Space Walker
Space Walker
User avatar
Joined: Fri Sep 10, 2004 5:35 am
Posts: 207
Location: New Zealand/Europe
Post Modular Rocket Design   Posted on: Sat Aug 01, 2009 10:59 am
Glancing over the IOS website I see they have opted for a modular rocket design now

http://interorbital.com/Neptune%20Modular%20Page_1.htm

I guess this would keep development costs down and enable the project to be fast tracked, has anyone done the math or something for the most optimum in multi engine designs? I mean the Neptune 4000 they state will have 48 booster engines, 24 stage II engines etc.. or is it more the case these days of having a reliable engine/setup and avionics package, so the number doesn't really count.
I'm just curious, was thinking back to the N-1 days when I was reading this.


Iain


Back to top
Profile WWW
Space Walker
Space Walker
User avatar
Joined: Thu May 07, 2009 2:40 am
Posts: 159
Location: Austin, Texas
Post Re: Modular Rocket Design   Posted on: Sun Aug 02, 2009 2:44 pm
The modular design has some advantages, but it seems like it makes the rocket very short and wide. This is less efficient than a long skinny rocket. The modular design duplicates a lot of extra hardware, which will make the rocket heavier. It also increases drag on the rocket.


Back to top
Profile WWW
Spaceflight Trainee
Spaceflight Trainee
avatar
Joined: Tue Jun 30, 2009 6:51 am
Posts: 21
Post Re: Modular Rocket Design   Posted on: Sat Aug 08, 2009 9:02 am
DaveHein wrote:
The modular design has some advantages, but it seems like it makes the rocket very short and wide. This is less efficient than a long skinny rocket. The modular design duplicates a lot of extra hardware, which will make the rocket heavier. It also increases drag on the rocket.


Hugely paralleled launcher designs of this type are unpractical, any person can deduce that simply clustering hundreds of simple & cheap sub-systems. Will create an very complicated and inefficient system , with a greatly increased potential for failure.

For example say you build a basic propulsion module using blow down and hypergolic propellants it still requires.
a/ Tank cylindrical structure
b/ At least 3 tank domes (top + bottom + intertank)
c/ Pluming to convey propellants to motor
d/ At least 2 control valves
e/ Numerous fittings to connect each propellant line to valves etc ( at least 4 )

This into account each subsystem would have at least 20 potential failure points. Now multiple this by say 38 and you have around 750 potential failure points in their small vehicle. Human error is a big factor here, as it only takes somebody not tightening just 1 of those few hundred bolts or pressure lines.

Inspection and testing requirements would be greatly multiplied over a single large stages requirements, instead of 2 tanks you now have to test and inspect 100's. Has this been modeled into the cost? What is the continuity in this vehicle if say 2 modules fail to ignite?, or 1 or 2 cant be throttled for steering etc. To work around this you need high margins of saftey, which = more weight thus more modules.


Back to top
Profile WWW
Spaceflight Participant
Spaceflight Participant
User avatar
Joined: Tue Jun 15, 2004 7:54 am
Posts: 94
Location: Dallas, TX
Post Re: Modular Rocket Design   Posted on: Tue Aug 11, 2009 5:09 pm
I would say that the jury is still out on whether truly massive modular clusters with hundreds of modules are workable, but dismissing small launchers with a few dozen modules as unworkable because of part count shows a lack of understanding of the relative complexities of the vehicles. Look at a Soyuz, with 20 big regeneratively cooled nozzles from turbopumped engines, lots of verniers, and actively pressurized tanks. A 49 module OTRAG-like system would have a fraction of the part count. Really.

As for tankage, linear weld inches is a metric of sorts, but it is not uncommon for small tanks to be cheaper by volume than very large tanks. Quantity matters a lot for manufacturing. Out of curiosity, does anyone know if any launch vehicle has ever failed due to a tank fabrication flaw? Compared to engines, tanks are a lot easier to test.

John Carmack


Back to top
Profile WWW
Spaceflight Trainee
Spaceflight Trainee
User avatar
Joined: Mon Jun 16, 2008 1:10 pm
Posts: 31
Location: United States
Post Re: Modular Rocket Design   Posted on: Tue Aug 11, 2009 7:03 pm
I've followed the world's space programs since I was a child in the 60's. I've never come across any report of a loss due to a tank failure.

Regarding tanks, I design pressure vessels for a living. Doing a hydrostatic test on them is dirt simple, and with large numbers of repetitive circular welds, a simple robotic welder can turn out tanks as fast as you can load the parts in a jig. In fact, you can use a couple of robots to load tubes and preformed heads from racks to a jig, the jig clamps up, the robots weld it up, another loads it to a holding station to cool off, then another drops it into a test rig, closes off the connections, fills it with water, tests it, then moves on. (sorry for the run on sentence) Once you have the process' bugs worked out, you could produce copies of the same tank faster than you could sell them.

_________________
James Clem
Chief Engineer
Epsilon Vee

Veni Vidi Orbis


Back to top
Profile WWW
Space Station Commander
Space Station Commander
avatar
Joined: Thu Oct 27, 2005 7:44 am
Posts: 707
Location: Haarlem, The Netherlands
Post Re: Modular Rocket Design   Posted on: Wed Aug 12, 2009 5:52 am
Quote:
any person can deduce that simply clustering hundreds of simple & cheap sub-systems. Will create an very complicated and inefficient system , with a greatly increased potential for failure.


Have you ever used Google? That is a pretty accurate description of their infrastructure. In fact, just about all high-performance computing systems work that way these days...

_________________
Say, can you feel the thunder in the air? Just like the moment ’fore it hits – then it’s everywhere
What is this spell we’re under, do you care? The might to rise above it is now within your sphere
Machinae Supremacy – Sid Icarus


Back to top
Profile
Spaceflight Trainee
Spaceflight Trainee
avatar
Joined: Tue Jun 30, 2009 6:51 am
Posts: 21
Post Re: Modular Rocket Design   Posted on: Fri Aug 14, 2009 4:42 am
John Carmack wrote:
I would say that the jury is still out on whether truly massive modular clusters with hundreds of modules are workable, but dismissing small launchers with a few dozen modules as unworkable because of part count shows a lack of understanding of the relative complexities of the vehicles. Look at a Soyuz, with 20 big regeneratively cooled nozzles from turbopumped engines, lots of verniers, and actively pressurized tanks. A 49 module OTRAG-like system would have a fraction of the part count. Really.

John Carmack


John

I did say not practical, not unworkable. ;-)

My knowledge of complexities of various vehicles is not the issue, a Soyuz vehicle is a large launch vehicle. The cost of testing, QA etc can be amortized in its operational cost, which are still high enough that it offers no solution to the idea of "cheap space launch".

If one is trying to greatly reduce cost, operational procedures, and above risk the most logical approach is the lowest system/ parts count to achieve the desired goal. Unless you want to increase your reliability levels through heavy redundancy (large margins of safety etc), and more stringent testing regimes. The first normally adds more weight & the second adds greatly to the manufacturing & operational cost. All of which run contrary to what is required to produce a reliable small low cost launch system.

From memory structural failure is implicated in around 18% of launch vehicle mishaps, and propulsion in around 68%. So why as a system designer would you want to play these odds by greatly increasing the number of these individual sub-systems into your design?.

There is multiple issues here also with successful operation of all of the individual rocket motors, whilst strapped in parallel. Such as variations in thrust, combustion instability & oscillations, and potential for detonation of a motor. Something that becomes a real concern as the number of individual propulsion systems increases. Every time the amount of motors is increased they would need to conduct a ground based static test of the full engine cluster, to analyze the effects of coupling so many motors.

So I am not dismissing it & saying it wont work all, I am saying it will most likely not work reliably in low cost fashion.


Back to top
Profile WWW
Spaceflight Trainee
Spaceflight Trainee
avatar
Joined: Tue Jun 30, 2009 6:51 am
Posts: 21
Post Re: Modular Rocket Design   Posted on: Fri Aug 14, 2009 6:54 am
JamesC. wrote:
I've followed the world's space programs since I was a child in the 60's. I've never come across any report of a loss due to a tank failure.


James,

You’re only focusing on my mention of the tanks as a modular component.

I am not trying to say a tank failure is the only concern, but that it is an engineered system. And as we all know all engineered systems have a potential failure rate as a percentage of units made. Is that or is that not a basic engineering premise that we all agree on?. Especially when you must take into account lower factors of safety required by rockets to get the mass fractions required for usable propellant tanks. Of course you can make a gun hard tank out of steel with a burst factor of 3 to 1, and ensure a very low potential of structural failure. But can you use this tank for a launch vehicle - no.

So for this concept to work you need light weight tanks with a low safety margin, which means materials, weld quality etc all become far more critical and more likely to fail. And by the way early Atlas and centaur rockets had numerous in flight & pad mishaps due to propellant tank structural failures. Not to mention vehicles that have failed due to plumbing, valve & interconnect issues, all areas that a highly modular vehicle with hugely replicated systems would have increased risk of. For instance the first Space X Falcon fell out of the sky because an engineer failed to correctly tighten one pressure fitting. Now get a team of engineers to correctly without ever making a mistake to fasten and test many hundreds of such fittings, and do it cheaply and reliably. Reliably enough that vehicle with more than 50 of these can be grouped together numerous times and operate without failure, do you see my point yet?. I am sure you understand that a tanks internal pressure is not the only variable it is exposed during its operational life span.

Having said this in looking at inter-orbitals web they state they are going to use composite tanks, this opens the door to even greater automation. Which is great, but I would be pretty confident there would be no advantage in producing 100's of small composite tanks, versus one big tank. The big tank would be far more efficient in terms of volume vs weight, and would use less material than 100's of smaller units. The production rate of smaller tanks would require a multi-spindle or separate multiple winding machines to equal the production rate of 1 larger tank, so money saved producing a smaller mandrel would most likely be negated by the need to have multiple mandrels. Plus each tank would have a manual setup operation and some level of operator intervention to get an end product, once again greatly multiplied over a larger simpler system.

I haven't even discussed here the systems required to bond the hundreds of tank modules together either.

Regardless of how well we make them - rocket motors are rocket motors & at least one out of a certain number is going to fail, and the failures are normally not benign. so if your lucky and you can build a small motor that only fails one out of 100 times, this means in a vehicle where you have 100 engines one is bound to fail most times.

The OTRAG concept I believe was a spark of original thinking for its time, but it is far out of step with capabilities that now see very large composite pressure vessels as a common place industrial item. As well as know how in the area of large ablative/ composite liquid fuel motors, which I feel negate any of the perceived advantages of hugely paralleled boosters.

This just dosent make good sense from the perspective of trying to greatly reduce the cost, and raise the reliability of a small launcher design. And only seems attractive from a macroscopic point of view when considering the individual modules as been cheap & simple.

Inter-orbital had something I believe with the 1 and 1/2 stage concept that was viable, and technically sound.


Back to top
Profile WWW
Spaceflight Trainee
Spaceflight Trainee
User avatar
Joined: Mon Jun 16, 2008 1:10 pm
Posts: 31
Location: United States
Post Re: Modular Rocket Design   Posted on: Fri Aug 14, 2009 3:32 pm
Percentage of product failure.

Assuming that a certain percentage of product produced will fail is not an engineering premise. It’s a management premise, one that is very old, and in this day and age, very wrong. If I design it right, and you build it right, using the right materials, proper weld procedures, and someone else operates it correctly, within a properly determined number of operational cycles, it will work every time. There are companies, Toyota for example, that manage to achieve zero defect production. It’s not easy to set up, but it can be done. Just because someone might not tighten a nut properly is not an acceptable reason to assume a percentage of failure. It is an acceptable reason to put measures in place to ensure that it doesn’t happen.

Safety Factors.

When I design something, I must either make a decision what factor of safety is necessary, or if working from, say, ASME Code, the decision regarding the minimum acceptable factor is made for me. Safety factors are there to cover potential events that you have no control over, such as a tank overpressure because the local temperature hit a new record high today. Safety factors for a flight weight tank will be lower than that for a chemical reactor vessel in order to keep the weight down. That lower safety factor does lower the capacity to handle uncontrollable events. It does not increase the potential for failure.

Criticality of materials and procedures:

The more critical something in a design is, the more expensive it is. For instance, if I want a tank wall that’s only 1/32 inch thick, the manufacturing of said wall will have very low tolerances for over and under thickness values. It will be expensive. Welding said material is not going to be easy, requiring more highly skilled labor and time. It will be expensive. Neither of these facts has anything to do with rates of failure. If it’s built right, it won’t fail. But it will cost you.

Composite materials:

I don’t yet really know much about these, so I won’t comment.

Multiple modules:

I haven’t run the numbers yet, but small (50 or so) numbers might be workable for a small, dedicated cubesat type launcher. The reason I don’t think huge numbers of modules are workable is the total dry masses of bundled modules will rapidly outstrip a single tank with a smaller number of engines. Even than, building larger modules so you don’t have to use huge numbers of them might well be workable. I do think multiple modules won’t work for my N-Prize project, simply because it makes the launcher wider, and drag forces weight heavily on such a small vehicle.

Where am I heading with this?

Manufacturing has changed drastically over the last century. Acceptable defects will drive you out of business. For true space access we must decrease cost and increase reliability. We need to start by setting aside old assumptions.

_________________
James Clem
Chief Engineer
Epsilon Vee

Veni Vidi Orbis


Back to top
Profile WWW
Spaceflight Trainee
Spaceflight Trainee
avatar
Joined: Tue Jun 30, 2009 6:51 am
Posts: 21
Post Re: Modular Rocket Design   Posted on: Sat Aug 15, 2009 5:13 am
James.

In a very long winded way you have just agreed with me.

My argument has been all along the extra procedures & quality control required to make such vehicle reliable, would not create a cheap product. Not that it cant done, but that it wont be cheap!

Of course QA can be assigned to minimize risk and increase quality, but it cost and that is why today's current launchers are so expensive. Because they have so many high performing, complicated, high tolerance parts and sub-assembles that need to be 100% defect free. Thats why when you go and buy a Mil-Spec certified component it cost 10 times what an industrial one does.


Back to top
Profile WWW
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 10 posts ] 
Moderator: IOSCEO
 

Who is online 

Users browsing this forum: No registered users and 3 guests


cron
© 2014 The International Space Fellowship, developed by Gabitasoft Interactive. All Rights Reserved.  Privacy Policy | Terms of Use