Jump to content

What is it with the weird staging of the ISRO GSLV?


wb99999999

Recommended Posts

Hi everyone! I've been reading about the curiously designed ISRO vehicles lately, and here's something that deeply confuse me about the GSLV launchers:

The GSLV is a 4 staged rocket (of a sort). It has 3 stages stacked in tandem and 4 strap-on boosters acting as stage 0. The thing that confuses me a great deal is how the 2 starting stages are arranged. It uses a solid S139 booster as the core which does 4.7 MN with 269s Isp, and the four strap-ons are liquid boosters which does 760 kN each with 262s Isp. So far so good for me, even the usage of a solid as the core is a bit odd, but I totally understand the design choice since it has a higher Isp and is adopted directly from PSLV along with the second stage (which justifies using the S139 as the core stage given that they don't need to develop a new interface to mate with the second stage.).

However, it is the burn time and staging order that gives me trouble: not only the liquid strap-on boosters DOES NOT separate from the core (still okay for me, the Russian Angara had considered a similar staging choice), it has a whole minute longer burn time than the core. It doesn't make any sense to me. From my understanding, the last bit of fuel in a stage is the "best" fuel, since the fuel are only pushing the higher stage and payload, without pushing on a lot of yet unused fuel. But for this vehicle, the last 60 second of the boosters' fuel is literally pushing an empty, heavy solid rocket casing. If one consider the liquid boosters probably have a higher fuel fraction than the solid core, it comes out even worse. An analog that came to my mind is an Ariane 5 that does not throw away the solid booster after they has burnt out...

The point is, judging from the numbers, it seems to me that it is so OVERWHELMINGLY beneficial to at least jettison at least something for the first stage, even at the cost of complexity and some perhaps costly redesign. In my opinion, at this point the staging has gone beyond sub-optimal and entering the realm of counter productivity: the performance that the vehicle will gain from a better staging would be very large and this is exactly what they did for GSLV Mk3. The Mk3 uses a Titan style staging by the way, where strap-on boosters are ignited on the ground, and the core stage is only ignited after the booster separation.

So it follows that there has to be some reason that forced ISRO to opt for such a design. Could any of you friendly folks give me some insight on this?

Link to comment
Share on other sites

The GSLV first stage is the same as the PSLV (S139 solid stage), and their second stage is very similar, both powered by Vikas 4 engines.

My guess is that using this architecture, large boosters were the easiest way to push the design to 2.5t GTO. Being limited in power (India currently only produces Vikas variants in that range of thrust), bigger boosters meant longer burn times. Using proven components meant a cheap launcher and therefore they could allow losing a bit of efficiency by arranging the staging as they did (it's actually more "not exploiting possible efficiency" that "losing efficiency" since the launcher fills its role which is enough).

Link to comment
Share on other sites

It may not be the most optimal design overall, but it is apparently the most optimal with the components they already have developed and available. Saves a lot of R&D time and money. Given more R&D time and funding, they could probably come up with a more powerful, longer burning core stage for more payload. In the interim, this'll do, just like STS and its Block 1 second stage, with plans for a newer, more powerful second stage to be developed.

Link to comment
Share on other sites

9 minutes ago, mikegarrison said:

Solid rocket engines aren't particularly heavy. Unlike liquids, there are no pumps or combustion chambers. They are basically just a tube with a nozzle stuck onto the end of it.

Precisely. Keeping it there gives a central thrust column, if nothing else.

Link to comment
Share on other sites

1 minute ago, sevenperforce said:

Precisely. Keeping it there gives a central thrust column, if nothing else.

There are tradeoffs. The tube that holds the fuel has to take higher pressures than a liquid rocket tank, so it can't be as light. Instead of a heavy engine, it has a heavy "fuel tank".

Anyway, I'm sure the relevant point here is one that is probably familiar to any KSP player. Instead of re-optimizing all the staging fresh, they just added "moar boosters".

Link to comment
Share on other sites

42 minutes ago, mikegarrison said:

There are tradeoffs. The tube that holds the fuel has to take higher pressures than a liquid rocket tank, so it can't be as light. Instead of a heavy engine, it has a heavy "fuel tank".

Anyway, I'm sure the relevant point here is one that is probably familiar to any KSP player. Instead of re-optimizing all the staging fresh, they just added "moar boosters".

 

Your point exactly. If I'm not getting something horribly wrong, the empty mass of a solid fuel booster would exceed a similarly-sized liquid rocket assembly for quite a bit. A liquid rocket's tank is just a thin layer of usually aluminium or stainless steel, while a solid's casing have to be a lot thicker and heavier to withstand the combustion pressure, acting effectively as the combustion chamber of a liquid fuel engine.

I probably played way too much RO and became alien to the idea of Moar Booster, but it is a valid point. ISRO is apparently a tightly budgeted agency...

Link to comment
Share on other sites

12 minutes ago, wb99999999 said:

Your point exactly. If I'm not getting something horribly wrong, the empty mass of a solid fuel booster would exceed a similarly-sized liquid rocket assembly for quite a bit. A liquid rocket's tank is just a thin layer of usually aluminium or stainless steel, while a solid's casing have to be a lot thicker and heavier to withstand the combustion pressure, acting effectively as the combustion chamber of a liquid fuel engine.

I probably played way too much RO and became alien to the idea of Moar Booster, but it is a valid point. ISRO is apparently a tightly budgeted agency...

Sevenper also had a valid point that you have to have some structure to carry the thrust loads anyway.

The main point is that this is the real world, and rockets are not actually optimized on their fuel fraction but rather on their cost-to-orbit. And cost includes things like designing a whole new rocket versus just strapping a few boosters on one that already exists.

Link to comment
Share on other sites

3 minutes ago, mikegarrison said:

Sevenper also had a valid point that you have to have some structure to carry the thrust loads anyway.

The main point is that this is the real world, and rockets are not actually optimized on their fuel fraction but rather on their cost-to-orbit. And cost includes things like designing a whole new rocket versus just strapping a few boosters on one that already exists.

I see. Still, it is how inoptimised the whole thing is that made me scratch my head.. I mean there're just so many compromises in the design, I'm not even sure the cost reduction and technological readiness is enough to justfy it...

This thing is just so counter-intuitive, with a longer burning stage actually having lower Isp but higher fuel fraction and non-separating booster and massive dead weight... the "cost per performance" seems very grim...

Link to comment
Share on other sites

As others have said, welcome to real-world engineering. There's no such thing as a sub-optimal design if it meets all of it's design goal and constraints, is within budget and also fits into current infrastructure, while also not requiring additional R&D costs. The only exception is if it's so badly designed that it is obviously under-performing compared to the potential of it's components, which I don't believe for one second this is.

True, the design is "sub-optimal" if you look at it from a pure maximum performance point of view, but I couldn't honestly tell you what changes you could make to remedy it without a complete redesign and concept change, and if you look at it from the point of view of the goals and capabilities that the ISRO, then its a perfectly good design.

EDIT: Let's also look at in in another way. Let's say they did completely redesign the first stage to get rid of some of the inefficiencies you've mentioned. Even if they found an extra half-tonne of payload to LEO (which I would imagine is wildly optimistic), where has all that time and money actually got them? How often are they going to use that extra capability when the rocket was designed to have the capability it has now (so logically ISRO thought that very few payloads would exceed it's current capability)?

Edited by Steel
Link to comment
Share on other sites

4 hours ago, wb99999999 said:

This thing is just so counter-intuitive, with a longer burning stage actually having lower Isp but higher fuel fraction and non-separating booster and massive dead weight... the "cost per performance" seems very grim...

Except the longer burning "boosters" have a better Isp than the solid "core" (254s sea-level, 287s vacuum for the Vikas, vs 237s sea-level, 269 vacuum for the S139)

Probabaly easier/cheaper to build this way with the engines available than to try to build an asymmetric design with one solid booster and a four engine core.

Link to comment
Share on other sites

44 minutes ago, Steel said:

 How often are they going to use that extra capability when the rocket was designed to have the capability it has now (so logically ISRO thought that very few payloads would exceed it's current capability)?

ISRO already produce many satellites that are too large for GSLV mk. 1/2, they just launch on Ariane currently. It's a cludge put together from existing components so they don't have to buy as many foreign launches, not something sized to launch anything they might want to.

 GSLV mk. 3 is being developed with nearly twice the capability, but ISRO have already developed sats too big for even it.

 

Link to comment
Share on other sites

9 hours ago, wb99999999 said:

Your point exactly. If I'm not getting something horribly wrong, the empty mass of a solid fuel booster would exceed a similarly-sized liquid rocket assembly for quite a bit. A liquid rocket's tank is just a thin layer of usually aluminium or stainless steel, while a solid's casing have to be a lot thicker and heavier to withstand the combustion pressure, acting effectively as the combustion chamber of a liquid fuel engine.

The empty solid booster is heavier than a liquid-propellant tank of equivalent volume, but it is lighter than the entire liquid rocket assembly would be.

Link to comment
Share on other sites

14 hours ago, wb99999999 said:

From my understanding, the last bit of fuel in a stage is the "best" fuel, since the fuel are only pushing the higher stage and payload, without pushing on a lot of yet unused fuel

It is entirely possible that the rocket can't support the extra force of the acceleration for all rockets firing right before they are empty.  I'd expect this with crewed vehicles, but there are always support issues with any rocket.

5 hours ago, Rhomphaia said:

Except the longer burning "boosters" have a better Isp than the solid "core" (254s sea-level, 287s vacuum for the Vikas, vs 237s sea-level, 269 vacuum for the S139)

Probabaly easier/cheaper to build this way with the engines available than to try to build an asymmetric design with one solid booster and a four engine core.

It should be easier to configure the solid booster for more thrust (to a limit, and of course your rocket has force limits as well), and if the Isp is less than burning it up first is better still.

Link to comment
Share on other sites

1 hour ago, sevenperforce said:

The empty solid booster is heavier than a liquid-propellant tank of equivalent volume, but it is lighter than the entire liquid rocket assembly would be.

GSLV first stage mass fraction is 82%, you'd be hard-pressed to find a liquid stage less efficient than that; the liquid core on GSLV mk. 3 gets 91%. 

Link to comment
Share on other sites

Because the liquid engines burn longer than the solid engine and are attached to the solid engine, the solid cannot be dropped first. Making the liquid engines burn quicker would require either adding more ($$$), making each produce more thrust ($$$) or reducing fuel load which absolutely reduces performance. Making the solid burn longer with the same fuel mass would reduce liftoff TWR, increasing gravity losses, and would also increase peak g-force which might be undesirable. And totally redesigning the structure is $$$.

I'd have thought giving the solid more fuel would have been relatively inexpensive though. Maybe it would affect the aerodynamics? Or maybe it's just not worth it?

In any case, "totally redesigning the structure" is exactly what's being done. The GSLV Mk 3 will have two S200 solid boosters for its first stage, and a core with two Vikas engines that ignites in-flight thus being functionally the second stage. That compares to the current GSLV with its four Vikas-engined 'boosters' and an S139 solid core, all ignited at launch.

Link to comment
Share on other sites

24 minutes ago, cantab said:

I'd have thought giving the solid more fuel would have been relatively inexpensive though. Maybe it would affect the aerodynamics? Or maybe it's just not worth it?

Just not worth it. It's a pre-existing stage already being built for their polar sat launcher. It was close enough to what they needed for their geostationary sat launcher, so they used it as is.

Why does the 737 of today use basically the same flight deck windows as the 707? It's because the design works well enough, and it would be expensive to redesign and then recertify a new window.

Edited by mikegarrison
Link to comment
Share on other sites

I decided to quickly play around in KSP and knocked up the various staging arrangements.

Solid core, liquid boosters (that burn for longer), all ignite at launch: 2682 m/s
Solid core, liquid boosters, core ignites after booster sep: 2721 m/s
Liquid core, solid boosters, all ignite at launch: 2928 m/s
Liquid core, solid boosters, core ignites after booster sep: 3106 m/s

Screenshots: https://flic.kr/s/aHskYQKeGs

So the seemingly-odd GSLV approach does not appear to be as detrimental as you might expect. It loses about 14% of the delta-V compared to doing things the 'right' way, which is noticeable but doesn't seem extreme.

Thinking about it, sure, so the 'boosters' are lifting the empty SRB core which is 30 tonnes of steel. But when the second stage has 42 tonnes of fuel, the third stage 12 tonnes, plus dry mass for everything (which I haven't found figure for), well ... actually that empty SRB does look kind of heavy percentage-wise after all.

Link to comment
Share on other sites

6 hours ago, cantab said:

Thinking about it, sure, so the 'boosters' are lifting the empty SRB core which is 30 tonnes of steel. But when the second stage has 42 tonnes of fuel, the third stage 12 tonnes, plus dry mass for everything (which I haven't found figure for), well ... actually that empty SRB does look kind of heavy percentage-wise after all.

http://www.b14643.de/Spacerockets_1/India/GSLV/Design/GSLV-2.htm has gross and propellant masses for all stages.

Link to comment
Share on other sites

I'd put my money on economics.

That is usually the answer when something does not match the engineering paradigm.

Possibly it was faster to construct it this way, maybe there was a deadline to meet? Slightly inefficient design made up for by meeting a contract?

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...