Which engine for the SRB launcher's upper stage

[dwightlooi]

It seems that NASA's choices are few if the engine is to be a high thrust LH2/LOX liquid fuel engine.<br /><br />(1) <b>J-2<br /></b><br />The Apollo era J-2 is a 200,000 lbs thrust gas generator engine. It has been proven in flight and is so called "man rated". But there is no J-2 production line and its performance is so-so by today's standards -- with an IpSec in the 420 to 430 range. In addition, a single J2 only makes about 91,000 kg of thrust. This limits the combined mass of the upper stage and payload to roughly 90,000 kg when the SRB will happily lift a total combined upper stage mass in excess of three times that and still have a lift off thrust to wieght ratio no worse than the Shuttle's. It is not a very good choice for maximizing payload performance.<br /><br />(2) <b>SSME<br /></b><br />This will certianly work. The SSME is a high efficiency, staged combustion engine. 213 ton worth of thrust means that a 200 ton class upper stage plus pay load combo can be used. Having practically twice the upper stage propellant will greatly enhance the payload capacity. Its IpSec of 455 secs is also very good. In fact, the SSME will perform better on an SRB launcher than on the shuttle since the engine is seriously over expanded at liftoff and midair lighting will allow it to perform better over its total burn duration. Unfortunately, this is also THE MOST expensive and complex engine available anywhere. Even if some cost savings can be had from not having to make the engine reusable, it is still going to be one costly powerplant costing as much as $20-40 million.<br /><br />(3) <b>RS-68</b><br /><br />The Delta IV's RS-68 is another gas generator engine. In theory that makes it a safer design than the SSME because the overall operating stress and complexity is lower. Thrust from this powerplant is really more than is needed. At 338 tons it will lift any amount of upper stage fuel and structure which the SRB can get off the launch pad. However, its efficiency is rather bad for a

 

[cuddlyrocket]

Of course, NASA could license Snecma's Vulcain engines, and manufacture them in the US. The Vulcain 1 has a thrust of 256,000 lb and an ISP of 431s, with the Vulcain 2 having 300,000 lb and 433s.<br /><br />The advantage is that it's cheaper and quicker than developing a new engine. The disadvantage is that it's non-American (un-American?) - even worse, it's French. But hey, it's your money.

 

[dwightlooi]

<i>Most seem to favor the J2S. Which was being developed from the J2.<br /><br />The SSME is not designed for an air start and is very expensive.<br /><br />The RS-68 is too heavy.</i><br /><br />I don't think the weight is an issue. First of all, every single one of those engines are relatively light in context of the mass of the upper stage and payload. The RS-68 is 6.6 tons, this is relatively insignificant in view of a 300 ton upper stage. In fact, the RS-68 has roughly the same thrust to weight ratio compared as the J2; the J2 makes 91 tons of thrust weighing 1580kg, whereas the RS-68 makes 365,000 kg of thrust weighing 6,600 kg. Quite remarkable for an engine designed for low parts count, minimal machining time and using an ablative nozzle. The biggest technical disadvantage is that its IpSec performance (410 sec) sucks dirt in the realm of LH2/LOX engines. The main reason, or excuse, NASA may not use it is that it is not "man-rated". But neither is a modified J2, since whatever modifications involved will require that it be re-rated for human space flight.<br /><br /><br /><i>Of course, NASA could license Snecma's Vulcain engines, and manufacture them in the US. The Vulcain 1 has a thrust of 256,000 lb and an ISP of 431s, with the Vulcain 2 having 300,000 lb and 433s.<br /><br />The advantage is that it's cheaper and quicker than developing a new engine. The disadvantage is that it's non-American (un-American?) - even worse, it's French. But hey, it's your money. </i><br /><br /><br />The nation of origin aside, the Vulcian does not provide a significant improvement over the J-2. IpSec is about the same and so is thrust. The Vulcain is neither as high performance as the SSME, nor is it as powerful or as cheap as the RS-68. Like the RS-68 it is not "man-rated" yet and that is NASA's favorite excuse.<br />

 

[drwayne]

The RL10 was/is a very capable, versatile engine.<br /><br />Wayne <div class="Discussion_UserSignature"> <p>"1) Give no quarter; 2) Take no prisoners; 3) Sink everything."  Admiral Jackie Fisher</p> </div>

 

[dwightlooi]

The RL is too weak for the SRB upper stage. Even the RL-10B2 used in the Delta IV -- the most advanced in the family -- has a mere 11,200 kg of thrust. That is not going to work for an upper stage that will light 180 sec into the flight and must lift between 100,000 kg and 300,000 kg (structure, payload anf fuel); not even in pairs or trios. A single J-2 alone makes 8 times the thrust, an SSME makes 20 times and an RS-68 30 times. You cannot cluster enough RL10s to make it feasible.

 

[mikejz]

Given that we do not know the weight of the CEV, might a non LH2 Engine suffice? Like the RS-27 (237,000lbs, weight 2.5 tons, ISP 302). Of course the ISP is about 100 less than the hydrogen alternatives, but seeing the thrust of the single stick, it might not be a big factor.

 

[drwayne]

I have seen some proposals over the years that would dramatically improve RL10 performance, but I don't have the numbers with me...I am not even sure where in this mess they are....what a putz...<br /><br />Wayne <div class="Discussion_UserSignature"> <p>"1) Give no quarter; 2) Take no prisoners; 3) Sink everything."  Admiral Jackie Fisher</p> </div>

 

[scottb50]

http://www.pratt-whitney.com/prod_space_rlx.asp <div class="Discussion_UserSignature"> </div>

 

[dwightlooi]

Lets lay out the basic numbers...<br /><br /><b>The SRB -- as it is built for use with the shuttle -- makes 1,175,000 kg of thrust at lift off and burns for 124 secs. It weighs 590,000 kg loaded and primed for launch.</b><br /><br /><b>If NASA uses the SRB as is (which I seriously doubt for reasons I'll explain later) there will be a 585,000 kg thrust surplus at lift off.</b> This gives us an absolute upper bound for the combined weight of the upper stage and payload. It must weigh in below 585,000 kg, otherwise the thing won't get off the pad. <b>Less obvious however is that the upper stage cannot be too light! If you put something that is say 100,000 kg on it, not only will the payload rating be small, the G-load expereinced will be in the neighborhood of 15Gs -- too much for any astronaut, gung ho or not.</b> Hence the upper stage must be sufficiently heavy to dampen the acceleration. Of course a larger, heavier upper stage also means more upper stage fuel and payload capacity. If the upper stage weighs 200,000 kg G loads, while still high, will be acceptable (less than 9Gs) and payload rating will be in the 25 ton class. However, this will require an engine which will produce at least 200,000 kg of thrust. Two Modified J2s may do it, one SSME may also do it. Things get even more interesting if NASA opts for a 300,000 kg upper stage. This will require an RS-68 class engine and will make the ascent G loadings very comfortable. It will also increase the payload capacity to around 30 tons.<br /><br /><b>Now, I said earlier that I do not think that NASA will use the SRBs "as is". This is because, the SRB's thrust profile is really far less than ideal.</b> It is designed to lift the heavy shuttle and tank. It makes too much thrust for too short a duration. <b>This means that NASA has to either use a very big upper stage or accept very high G-loads.</b> Simple math shows that the same SRB, with the same IpSec, but configured to burn more slowly will be a better ideal. <b>If</b>

 

[najab]

I imagine that a regrained SRB might be used for the second-generation CEV. I don't see it happening for the first spiral.

 

[dwightlooi]

It is not that hard to "regrain" the SRB, especially if the aim is to lower thrust not increase it. The nozzle, gimballing mechanism, casings, etc remain the same. There will be no over pressure qualification needed since pressure is going to be lowered. The Propellant is probably going to be the same too. The only thing that changes is probably the internal geometry of the propellant. Probably a less pronounced "star" pattern in the central channel so as to lower burn rate. This is very easily done on the manufacturing side and very easily modeled on a computer.

 

[najab]

I don't know how easy it would be on the manufacturing side. The current pour is <b>very</b> exacting and there was a long qualification process required - changing the process is not something I'd classify as easy.<p>In theory, the procedures could be relaxed somewhat since there isn't the need to have two boosters with the exact same thrust profile, but it still wouldn't be a minor change.</p>

 

[gunsandrockets]

I prefer a cluster of five P&W RL-60 for the second stage. ISP of 465 and combined thrust over 250,000 lbs. No development needed unlike a modernized J-2S or an airstart SSME, plus the cluster permits engine out capability.

 

[dwightlooi]

<i>I prefer a cluster of five P&W RL-60 for the second stage. ISP of 465 and combined thrust over 250,000 lbs. No development needed unlike a modernized J-2S or an airstart SSME, plus the cluster permits engine out capability.</i><br /><br />The RL-60 is a berand new engine. It has done quite a bit of ground testing, but it hasn't flown on anything yet. It is also not manrated by any NASA definition -- I happen to believe that the whole manrating business is a load of bull, but I am sure NASA disagrees. Hence to say that the RL-60 does not need any further development is an understatement.<br /><br />I don't think clustering 5 RL60s is feasible in anycase. Firstly, I doubt five of them will fit in the base of a 4 or 5m upper stage. Secondly, I don't think they will be any cheaper than a single SSME by the time you multiply its unit cost by five. <br /><br />If the preference is for a 250,000 + lbs of thrust and an inherently safe expander cycle engine, I think NASA will be better off finishing up on the development of the 300,000 lbs RLX and using that.

 

[frodo1008]

I wonder how P&W's (actually United Technologies, P&W's parent company) recent purchase of Rocketdyne from Boeing will affect these and other large liquid rocket engines? <br /><br />I have read where the merger is a good mix for United Technologies as P&W has generally concentrated on upper stage lower thrust engines, while Rocketdyne is the chief maker of the more powerful lower stage engines in at least the United States, if not the free world. <br /><br />I know that as a retiree of Rocketdyne while it was still under Boeing control I am at least a little prejudiced here, but I really think that this will prove to be a very bad decision on the part of Boeing, and a very good one for Untied Technologies, in particular in the long run!!<br /><br />Knowing both the capabilities, facilities, and even at least some of the people of Rocketdyne I Can tell you that at some $700 million in cash United Technologies got one of the best bargains ever to be had in the world of aerospace!!

 

[dwightlooi]

Do you believe that, if given the funding and the job, Rocketdyne or P&W can get the RLX engine finished by 2010?

 

[tap_sa]

<font color="yellow">"The RL-60 is a berand new engine. <b>It has done quite a bit of ground testing</b>, but it hasn't flown on anything yet. "</font><br /><br />Are you sure about this, got any links? Pretty much every trail on internet ends to news in summer 2003 how it's 90% ready and groundtesting starting in September. No news anywhere after that. AFAIK both RL60 and RLX got most of their funding from SLI, which got pretty much axed?<br /><br />S_G's 1500 tonne initial SRB thrust news is very good, otherwise the lift-off acceleration would have been very sluggish. Crude rule of thumb; 2nd stage wet weight = 10 x payload. It's still unclear how much CEV will weigh, isn't it? If it's the lifting body then probably closer to 30t, otherwise less.

 

[gunsandrockets]

"I don't think clustering 5 RL60s is feasible in anycase. Firstly, I doubt five of them will fit in the base of a 4 or 5m upper stage. "<br /><br />Well the Falcon V rocket will fit five 100,000 lb thrust Merlin engines on a stage 3.65 meters in diameter.<br /><br />http://www.spacex.com/<br /><br />The old Saturn IV fit six RL-10 engines on a stage 5.5 meters in diameter. <br /><br />http://www.astronautix.com/stages/saturniv.htm<br /><br />So clustering five RL-60 on an upper stage shouldn't be a problem.<br /><br /><br />" Secondly, I don't think they [RL-60 cluster] will be any cheaper than a single SSME by the time you multiply its [RL-60] unit cost by five... I think NASA will be better off finishing up on the development of the 300,000 lbs RLX and using that. "<br /><br />Perhaps. I doubt the cost difference would be significant considering both the SSME and the RLX are designed for reuse (up to 100 times in the case of the RLX) whereas the RL-60 is an expendable design.<br /><br />"The RL-60 is a brand new engine. It has done quite a bit of ground testing, but it hasn't flown on anything yet."<br /><br />That still puts the RL-60 way ahead of the not even tested yet airstart SSME, updated J-2S or RLX. Plus the RL-60 has an ISP superior to any of those engines.<br /><br /><br />

 

[krrr]

"If you put something that is say 100,000 kg on it, not only will the payload rating be small, the G-load expereinced will be in the neighborhood of 15Gs."<br /><br />You're overestimating the G-load here, probably by not taking the SRB's empty mass (86 tonnes) into account. Assuming a near-burnout thrust of 1400 tonnes (it's likely less), upper stage plus payload mass of 100 tonnes results in 7.5 G. For 150 tons, it's 6 G, for 200 tonnes, 4.9 G.<br /><br />Also, I don't think the upper stage's thrust must necessarily be the same or more than stage-plus-payload mass. For instance, the Ariane 5G still weights around 150 tonnes after booster separation, with a thrust of 113 tonnes.<br /><br />Therefore, at the low end, a 120 tonnes stage with a single J-2S might be feasible. Incidentially, that's a Saturn V third stage. Payload is of course modest, maybe 12 to 14 tonnes to an ISS orbit.

 

[gunsandrockets]

"The cluster makes an engine out 6 times as likly. "<br /><br />So? If an upper stage with a single SSME suffers an engine failure, it results in mission failure. Wheras an upper stage with engine out capability can still complete the mission even if one of it's engines fail.<br /><br />SpaceX is quite proud of the five engine design of the Falcon V launcher because they claim it improves mission reliability...<br /><br />"The major breakthrough in reliability is derived from the five engine, single tank first stage configuration. As such, Falcon V is the first American launch vehicle since the Saturn V to offer true engine out reliability. Moreover, depending on the phase of flight, the first stage can lose as many as three engines and still complete its mission."<br /><br />"A simple analysis, assuming an equal 0.5% chance of a given engine failing, means that Falcon V has a factor of ten higher propulsion reliability than another launch vehicle with one engine. There is a 5 * 0.5% chance of one engine failing, multiplied by a 4 * 0.5% chance of a second engine failing, which gives a 0.05% chance of propulsion causing mission failure. The Falcon V advantage is actually greater in reality, because it can lose more than one engine and safely complete its mission."

 

[shoogerbrugge]

spaceX uses 5 engines on its Falcon V because it wants common parts, and can't afford to yet design a different engine. <br /><br />the SpaceX huff and puff about reliability is moot point if they haven't launched a bit. Trackrecord is important, and a good trackrecord can be achieved with with uncomplicated rockets (Atlas V) and extremly complicated machines such as Proton.<br /><br />

 

[krrr]

This article gives a maximum G of 3.42 during first-stage burn, which translates to less than 800,000 kgf at SRB burnout. Burn time is still given as 124.6 seconds.<br /><br />So either the SRB thrust profile is different from what has been said here, or they will indeed have to "regrain" it.

 

[frodo1008]

I know that anyone who actually has some kind of experience, and gives the slightest bit of skepticism is immediately labeled as some kind of jerk by most of the supporters of this project (the Falcon series by spacex). However, I really think that some kind of reality thinking here is needed. Elon Musk himself was somewhat surprised at the technical difficulties of even the Falcon I level of vehicle, so he at least deserves some kind of respect for honesty. Also, this company is saying that it will be developing these rockets in a relatively hostile current launch vehicle market. <br /><br />I do think that the Falcon I will indeed be quite successful as the market for satellite launchers below 1,000 lbs to LEO is somewhat wide open. However, the market for larger launchers is very tight, and likely to get even tighter. I am not just talking about Delta II, and the original Atlas designs. There is also Arian space, the Russians, and the up and coming Chinese. All have good rockets in the general satellite launching business.<br /><br />Then there is a problem that those who have never dealt with cutting metal for these rockets and their engines do not seem to know about. The sheer increase in diameter size to build even a relatively small Delta II rocket system necessitates the buying of far larger and more expensive machining equipment. When we had to machine the 8 foot in diameter aft manifold ring of the SSME at Rocketdyne, we needed a Vertical Turret Lathe of some 10 foot capacity. Such machines are $millions of dollars worth of capital investment alone. Even the material handling equipment for such projects (cranes, fork lifts, etc, etc) become major investments alone. By the way the much hated regular aerospace companies like Boeing and LM already have a whole lot of such equipment available. So IF they really thought they were truly threatened by such operations as space-x they could also lower their respective costs! Which would indeed alone just

 

[krrr]

OK, I finally found a diagram of the SRB thrust profile in the CAIB Working Scenario.<br /><br />Basically, thrust starts at 1270 tonnes, builds up to 1400 tonnes at 22 seconds, drops to 1020 tonnes at 54 seconds, rises again to 1160 tonnes at 78 seconds, then falls to 726 tonnes at 110 seconds and then even more steeply until burnout.<br /><br />Conclusion: The thrust profile is already friendly enough for an SRB-based CEV launcher.

 

[gunsandrockets]

Thanx for that detailed information on the SRB thrust.