Northstar1989

Reproduced the crash from before, more or less. Was able to swap successfully back to the plane and land it, but when choosing to return to the main mission, the game crashed. A Dropbox link is attached below: https://www.dropbox.com/sh/qnndb0yy6iygw62/AABgS4jzfbq9_fWtvQekf4Jia?dl=0 You ran into no issues? That's interesting, but not my experiences. I keep getting crashes. I'm hoping it's not something stupid like overtaxing my limited RAM or free hardrive space on my main (C://) drive (even though the game is installed on D://) Anyways, the crash file is over 40 MB (created with Debug lvl 1 and 2 in FMRS enabled), so you should have plenty to work with there... Also, I didn't have too many problems with the window in any build so far- aside from it occasionally failing to minimize. Regards, Northstar

First of all Shynung, the tailplane is always aligned parallel to the main wing based on the way it is mounted, so any contradictory roll effects between the two wings is impossible, short of highly-localized lift imbalances that affect the front and rear wings very differently... And, while torques are possible if lift imbalances occur between the two sides of the wing, any torques placed on the tailplane are shared with the main wing- meaning the total level of necessary structural reinforcement required to deal with roll torques between the two remains more or less constant (not to mention it eliminates any possibility of the fuselages pitching differently, which can place some fairly serious stresses on the main wing attachments...) Structurally-speaking, you go from two weights suspended below a level board, to a box-shape, which is much stronger (especially in resisting any unequal torques in the yaw direction) and requires less structural mass for some of the same reasons that biplane wings were stronger than early monoplane wings, and allowed greater total wingspan and wing-loading... There's also the secondary or even tertiary advantage that merging the tailplane puts more Lift in the rear of the plane, meaning mass towards the rear of the plane is suspended from a closer point, and the fuselage needs less structural reinforcement to avoid sagging like a wet noodle... Ultimately it wouldn't matter what I say Shynung. I could say independent tailplanes were better and you would say they should be merged. I could state that straight wings had an advantage and you would insist on arguing in favor of swept wings. You're just arguing against whatever I suggest, and to a lesser degree in favor of whatever the current design utilizes, regardless of the facts. I find it quite disrespectful, and wish you would stop. Regards, Northstar

Don't select a mish-mash of out-of-context quotes, misrepresent that context, and think you have a point. You don't. For starters, I dropped any argument that the Stratolaunch flies transonic some time ago...You'll also have to cite that claim that the P-38 had problems at Mach 0.68 (as in an actual Wikipedia page/paragraph if it cones from Wiki), because that is a completely nonsensical statement. Teansonic regime starts *nowhere near* Mach 0.68, and the known aero problems of the P-38 were mainly in steep transonic dives...

Transonic performance isn't what we"re discussing here, so there's no need to bring that up. There was nothing in the P-38 design that called for twin fuselages- that's actually completely factually incorrect. In fact I just watched a 1-hour documentary on the P-38 last night that extensively discussed the reason it was designed as it was... The design requirements called for a specific top speed and altitude-ceiling, and Lockheed determined the only way to achieve the desired requirements was to use two engines. Having two fuselages to hold those two propellers just evolved naturally from the design requirements as a result. As for the connected tailplanes, it provided extra structural strength on the P-38 just as it would on the Stratolaunch 351. Today's matetials may be stronger, but that's no excuse to avoid making use of optimal design solutions where possible to increase structural strengthband allow use of less/thinner structural reinforcement to bring down structural mass... Having extra space to attach control surfaces doesn't mean that designers have to do so- but the same control surface attached further towards the rear of the plane has a longer lever-arm to the Center of Geavity, and thus you can get away with less total control-surface area. This reduces drag, mass, and cost- so it's certainly not a bad thing even with civilian aircraft... Simply saying the engineers are "smarter than us" are how we ended up with the incredibly-inefficient Shuttle. It's how we end up with drugs thst kill people and medical treatments that don't work (to name something closer to my own area of expertise- Biology). The lay public can and should be interested in, and yes even curiously critical of, scientific and engineering issues rather than professing ignorance about them- it's in everyone's best interest, because even engineers get lazy, stubborn, have ulterior motives, or make mistakes... I wasn't discussing the swept wings right at this moment, so your answer is completely off-mark. Though I don't think you were writing with this in kind, just in case, how would connecting the tailplanes together require any rework of the fuselages? They already have horizontal wing sections protruding from each tail boom- essentially all they would have to do is increase their span in the design until they had one single tail-wing (which should be stronger and simpler for the same reasons straight wings are compared to swept ones in this design) which they could slot through both tail booms instead of two seperate wings.

I've got it installed now, so I'll be testing it shortly. I've just been working on improving my Stratolaunch-style air-launch platform (swapping out some of the Goliath engines for ramjets, as they produce more Thrust at the relevant altitudes even at subsonic speeds, and increasing the wingspan even further- the point is to fly as high and at as low a dynamic pressure as possible to allow for the most stable and useful regime for decoupling the rocket payload...) while I waited to see if you learned anything useful from those crash logs... Regards, Northstar

The Stratolaunch doesn't fly supersonic, not anywhere close- in fact the entire past 3 pages of discussion have been largely about how fast it DOES fly, because that determines whether the aerodynamic advantage from swept wings is worth the structural stability issues they produce... The P-38 flew beautifully in subsonic regimes- so much so, in fact, that it was probably the most deadly mainstream (non-experimental, so early jets don't count) fighter in WW2. The top 3 American "Ace" pilots in WW2 all flew P-38's. Aerodynamics don't miraculously just change because a few decades have passed- and there is no reason to think connecting the tailplanes together on the Stratolaunch 351 wouldn't provide the same structural and aerodynamic stability benefits that doing so provided the P-38, as well as improved Lift:Drag ratio by raising the average Aspect Ratio of the plane's wings... Unlike the main wings, the tailplane wing sections would be relatively short, and so don't face the same limitations from the structural mass needing to increase so much along the rest of the wing length (to counteract bending) as to make extra wingspan no longer beneficial. And if increasing the tailplane wingspan and connecting the tailplanes together made the Stratolaunch 351 *too stable* they could always move the main wings back a little to compensate (which would also have the useful side-benefits of reducing the stresses on the fuselages, and reducing the shift in Center of Gravity when the payload is released...) Once again, here are the images of the Roc and P-38 so everybody can see and try to visualize what I'm talking about... Stratolaunch 351: https://goo.gl/images/rwK0St P-38 "Lightning" https://goo.gl/images/OxSSXV Regards, Northstar

That's a good question- and why I started this thread: to try and figure out why Scaled Composites designed the Roc the way they did... The more I look at it, the more I am starting to think they just took earlier designs like that of the White Knight carrier plane- and just scaled them up. Otherwise, if they were going for an efficient airframe layout, why not connect the tailplanes together P-38 style, for instance? A simple comparison of it and the P-38 will show those tail booms are just BEGGING to be connected. Doing so would also allow them to REDUCE the chord of their main wing a bit while maintaining the same overall wing-area: leading to a higher Aspect Ratio and thus more Lift for the same Drag- as well as the structural strength benefits doing this would convey... Stratolaunch 351: https://goo.gl/images/rwK0St P-38 "Lightning" https://goo.gl/images/OxSSXV Let's also not forget that moving more of the wing to the rear like that would also make the Stratolaunch 351 more inherently aerodynamically stable- and provide a large area to attach elevators to the rear edge of the tailplane... Regards, Northstar

OK, so this time FMRS managed to successfully load up one of the dropped stages (my plane) without crashing- but instead crashed when I tried to swap back to the main vessel (the rocket) after landing the dropped stage. I was running FMRS on Debug levels 1 and 2, so in addition to the usual error.log file it produced a massive 56 MB output_log.txt file Since that's much too large to copy-and-paste here, I have included a Dropbox link to the files below: https://www.dropbox.com/sh/wv1s6gzgpj11fng/AACCdx9gdV7AlGGCWNHD7Vroa?dl=0 No idea what went wrong here- but I know FMRS used to work fine for me on this laptop on previous versions of KSP. I'm hoping you can sort through the issue here. I didn't get to this before re-creating the crash issue, so the files I posted correspond to the previous debug version (1.03) you posted here before. Regards, Northstar

The fact that it's already carrying a draggy payload reduces the penalty of the wing chord. Since the Lift:Drag ratio of the plane as a whole is determined by adding up a weighted average of ALL the components of the design, having an already low average means that there is less of a penalty for lowering the Lift:Drag ratio of the main wing. Indeed, by increasing the absolute magnitude of the wing's Lift and Drag doing so may actually INCREASE the plane's overall Lift:Drag ratio (similar to this problem- which is greater: "9+4+1+2 / 4" or "7+7+4+1+2 / 5" ?) What adding extra engines and wing chord actually does is increase excess power at a given altitude and speed. This power can then be divided up among a number of competing priorities, including payload, structure (safety/manufacturing margins), altitude-ceiling and climb rate (basically, not adding any extra mass and instead letting the plane climb higher), or fuel-capacity. And it's possible to split the difference between one or more of these, for instance adding a bit of fuel and a bit of payload, or reinforcing the structure a bit but not enough to prevent an overall increase in altitude-ceiling... The Stratolaunch 351 design has, according to information provided by its designers to PopularMechanics, an intended round-trip range of 1500 miles (that is, it can fly 1500 miles out to sea, release its payload, and then fly/glide without the rocket to a runway 1500 miles back). Which is already quite impressive, considering the size/weight if its payload- and, I would argue, a bit higher than necessary. It should certainly be possible to trade off some of that range capability for extra payload capability or better safety margins (low safety and manufacturing margins are one of the primary drivers of aircraft construction and maintenance costs, after all. When the aircraft parts have to be built to very high precision for the design as a whole to function, this adds a lot to the costs if production and requires more frequent maintenancw to maintain the aircraft within those narrow limits...) through an increase in the number of engines from 6 to 8 and an increase in the wing-area (mostly through wing-chord, as I assume they've already reached the practical/material limits for wingspan) as well. It would drive up fuel-costs to be sure, but in anything to do with space exploration fuel is only a tiny fraction of the overall cost of a mission... Finally, note that payload, range, and altitude-ceiling are, to some degree, interchangeable. That is, the Stratolaunch 351 could launch with a heavier payload on some days and climb to a lower altitude, and on others launch with a much lighter payload but launch it from a substantially higher altitude or further downrange. Thus, the extra power that adding engines and wing-area provides should in almost all cases be beneficial, and allow the Roc to carry a wider range of payload masses while benefitting all payloads... Regards, Northstar

Thanks. I managed to rebuild the plane/rocket pair so the root part was in the rocket by designing the rocket as a separate vessel and then merging the designs. Now FMRS recognizes the staging events correctly and designates the vessel I want it to as the Main Vessel by default (although it still would be nice if re-designating in the first 10 seconds after staging still worked). However, while FMRS recognizes each staging event, and correctly queues up each stage to switch back to, actually trying to DO so and switch back to a dropped stage, crashes the game instead- even though I waited to try to swap back to anything until the payload (a tourist in a Mk1 capsule, flown by a probe core) was already in orbit. Since the ability to pilot dropped stages back to the ground is the main functionality of FMRS, this is obviously a pretty serious bug. I only encountered it late last night/ this morning after I finally re-designated the root part so I wasn't trying to switch back to the stage I was currently piloting, and I had both debug levels toggled off at the time to minimize lag and try and ensure that running them caused no issues, but will try to find the time/motivation to reproduce the issue with debug levels both on at some point today... Regards, Northstar

Because I don't have the necessary skills to mess around with DLL's. Just look at my Mass Driver mod-fork (linked in my signature)- I haven't so much forgotten about it as reached a point where further updating it is beyond my (non-existent) coding ability, and was unable to find anyone willing to make necessary changes for me... Not meant to be mean, only to-the-point. Would a smiley have helped soften the request? Speaking of bugs, issues, and such- it appears the function to switch which stage is the "main vessel" within the first 10 seconds of a staging event isn't working on the 1.03 debug version fork. This isn't necessarily an issue for straight rockets- or even shuttles where the root part is located in the shuttle itself- but it *is* an issue for my attempt to carry out a Stratolaunch-style mission, where the root part lies in the plane but it's the detachable rocket that I need to fly to orbit and the plane that needs to land. Based on how the plane is constructed (with extensive use of surface attachments) any attempt to re-root so the parent part lies in the rocket crashes my game (this is a known issue with certain types of vessel builds in stock KSP), so I really need the function to work as intended so that I can fly the *rocket* to orbit and then switch back to the plane to land it (rather than FMRS thinking I want to do it vise-versa, and making the plane the "Main Vessel" at staging). Regards, Northstar

Why don"t you make these chsnges yourself, so the rest of us can enjoy a more functional version of FMRS until the original author returns to updating it?

Nobody knows how fast this plane is designed to fly, so your criticisms are baseless. <snip> Regardless of the speed the plane is designed to fly at, it's designed for altitude and payload, not fuel-efficiency, so one also wonders why they don't increase the wing chord (which would hurt subsonic Lift:Drag ratio but increase total Lift) and add the remaining 2 engines they have to the design. The 747's actually *FLEW* to the Mojave location, so presumably their engines were all functional- and if they added additional engines and wing chord, they could attain additional surplus power at the same max speed and altitude- meaning either a higher climb-rate and altitude-ceiling, or a higher payload-capacity (or alternatively, more generous saftey and manufacturing margins, and a cheaper, more rugged/reliable airframe), in all cases at the expense of range... Regards, Northstar

Flutter concerns are addressed by reinforcing the wing (which is heavier- but we already discussed this). Gliders and drones actually fly LOW, in relative terms- so you're defeating your own point there. And the Roc doesn't have a wet wing, as far as I know. Anyways, YNM already suggested a credible explanation regarding the ability to run the wing through the fuselages with a straight wing, but not with a swept wing- and since the Roc needs to carty an incredibly heavy payload below its center section, that actually makes a lot of sense (it would be much easier to carry the payload below a wing that ran through the fuselages than one thst was anchored externally). So I'm going to drop this here (provided nobody else posts anything I feel the need to reply to), as it looks like I have a good answer from YNM.

Since certain individuals keep insisting that raking the wingtips *reduces* the effective aspect-ratio if a wing, I *STRONGLY* recommend reading the Wikipedia section on raked wingtips. Raking the wingtips of a plane in fact *increases* effective aspect ratio, which is why raked wingtips are used: https://en.m.wikipedia.org/wiki/Wingtip_device So the main reason for the straight wings is so you can run the wings right through the two fuselages and treat it as a single section, structurally speaking? If that's even possible with this design (to run the wing through the two fuselages of the Roc), then I buy that. That *WOULD* pose a fairly large structural strength penalty on swept wings, far in excess of what the swept wings alone might pose. And considering they're already making the wings out if fairly flimsy carbon fiber panels... I guess that answers my question- as that's actually a highly plausible reason for the use of straight wings despite their inherent aerodynamic inefficiency (stronger wingtip vortices, reduced effective aspect ratio vs. raked wingtips, greatly reduced transonic stability, and a further-forward Center of Lift).

Spanwise flow is actually a GOOD thing in certain circumstances- it reduces drag in transonic conditions- it's disingenuous to list it as a negative. It DOES lead to vortices becoming a problem if you don't design the wingtips properly, but in fact the entire reason wingtips are often raked is to prevent this from becoming a problem by moving the vortices away from the main wing, and rectangular non-swept wings experience STRONGER vortices than wings with swept wingtips... Dutch roll could of course be a problem. But do please elaborate on how sweeping the wings back would lead to it, as it's one of the concepts I've always struggled to wrap my head around the causes of, and my plane designs in FAR actually often struggle not to suffer from it... "Lower effective aspect ratio" is also a disingenuous answer. If you proportionally reduce wing chord such that the wing-area of the wingtips remains unchanged then it's exactly the same... So, two bad answers that are outright wrong (aspect ratio *doesn't* decrease if wing area and span are held constant, and wingtip vortices caused by spanwise flow are actually STRONGER with straight, rectangular wings than tapered ones) and one- about Dutch Roll- that I don't understand as well as I would like to and would like to know more about before I make up my mind about one way or another, in short...

Such as? A few of you keep repeating that mantra, but provide absolutely no specifics to back it up. It"s not true *just* because somebody said it earlier. Runway clearance is a non-issue with the wings of the Stratolaunch 351 as the wings are already positioned so far above (nearly 40 feet!) the landing gear. And the extra mass required to structurally reinforce the wings to resist the extra bending moment created by sweeping the wings is only one (relatively minor) isdue, and has been discussed ad nauseum...

30,000 feet isn't high compared to any of those aircraft. And one cannot simply "design wings to provide more lift". Wings aren't some kind of magic where you wave a wand and get Lift- you have to purchase it with Drag. True, you can design higher aspect-ratio wings to provide better Lift:Drag ratio, but the only way to improve aspect ratio without compromising payload capacity is to increase wingspan. Since their wingspan is already at the limits of sanity (greater than any other aircraft in history) that's not really an option for a number of reasons including ground facility limitations and the need to make stronger/heavier wings to resist bending if they make them longer... Thus the logical way to gain additional altitude-ceiling for payload deployment would seem to be to add additional engines so they fly faster, and reduce Drag by sweeping/raking the wingtips back (which will also improve aerodynamic stability). They already do this with the wingtips on some commercial jets, there's no reason for you to treat it as something so extraordinary... For that matter, since they ought to be optimizing for payload and altitude, it would make perfect sense to just add on the 2 extra 747 engines they have access to and increase the wing chord if they're really going to fly slow enough that wing sweep doesn't matter. Decreasing the aspect ratio like this by thickening the wing chord would of course hurt the Lift:Drag ratio, but it would also provide more raw Lift for any given speed and altitude- meaning that with the extra Thrust the additional 2 engines would provide, they could "purchase" enough extra Lift to significantly improve the overall altitude-ceiling at the expense of fuel-efficiency and range... In the final analysis, having a big, heavy plane that's over-built is a good thing. It means that they can build and maintain the aircraft to looser engineering margins without increased risk of failure, bringing down overall costs. Given the purpose of the Stratolaunch 351 to merely serve as a launch platform, they should probably be aiming for big, rugged, and overbuilt- as really all the Roc needs to do is act as a super-heavy cargo plane of a sort (except one that releases its cargo midair...) Further, the heavier the plane is the more easily it should remain aerodynamically stable after releasing its rocket, as the rocket will be relatively lighter and its release will cause less overall shift in the Center of Mass... Thus the extra mass required to reinforce the wingtips enough to rake them back without causing excessive bending really wouldn't be such a bad thing. And it would actually *increase* maximum takeoff weight by more than the extra mass required, just like it does with commercial jetliners, if they stuck on enough 747 engines to ensure it could fly in a comparable speed range... Quit it Shymung. I have asked you politely now several times- drop that Straw Man argument right there. I never said anything, anything at all about the reason for the extra speed being to decrease the speed increase needed to attain orbit. I have in fact repeatedly said just the opposite- that the reason for it would be to allow the Roc to climb to a higher *altitude* before releasing its payload, and no other reason.

So... I'm getting all sorts of weird issues with the wings WITHOUT any version of FAR installed. Like the game treating the Center of Lift like it's at the attachment-point for the wings, and ignoring any wing-sweep that should be offsetting the CoL forwards or back. Is this a known issue? Is FAR required for the wings to function properly? I read the OP as meaning only that a certain LATER VERSION of FAR is required *IF* you use FAR... I've played with FAR in the past, but I'm trying to simulate a spaceplane based on a supersonic biplane airfoil- which in real life experiences LESS drag than a monoplane with comparable wing-area. However since FAR is programmed to ALWAYS penalize biplanes and doesn't have any way to simulate shockwave destructive-interference, the only way for me to get improved performance vs. a monoplane is to play stock... (where it at least gets improved Lift instead of reduced Drag) Plus, FAR overtaxes my current potato-laptop with its more advanced aero calculations. So I'd really like to avoid using it, and still have functional B9 swept wings if possible... Regards, Northstar P.S. I built the same basic wing shape out of stock parts before I replaced it with B9 Proc Wing parts- and it had none of the later design's enormous stability problems, and the CoM appears *well* ahead of the CoL in the Spaceplane Hanger and in pretty much the same place as in the stock version. So I know the issue usn't the wing shape or plane design, but how the game physics engine handles large, highly-swept B9 Proc Wings...

Don't put words in my mouth. I've repeatedly said that the benefit of the additional speed would be in allowing the carrier plane to fly higher- so it could release its payload at, say, 36000 feet instead of 30000 feet. 0.2 Mach is a roughly 30% increase in speed from Mach 0.66 to 0.86, so that represents a significant increase in lift for the same altitude. Saying I've said *anything* about the speed being important in and of itself is incredibly misleading, and I don't appreciate the Straw Man argument... Anyways, the entire PURPOSE of the carrier plane is to get the rocket as high as possible before release. Altitude makes a significant difference because it improves the ISP of the rocket engines and allows them to start firing more horizontally sooner in the ascent to orbit, and 6000 feet, for example, would represent a 20% in release altitude...

Lift increases with airspeed. That's a big part of why supersonic aircraft can fly so much higher than their subsonic counterparts with similar airframes- for example the Concorde flew much higher than the 747. The Stratolaunch 351 is supposedly only going to fly up to 30,000 feet, whereas the 747-100 could reach over 45,000 in level flight- so I don't see why they couldn't add on their two extra engines so they can climb higher/faster, and sweep their wingtips to reduce the drag. It would also make just getting off the runway easier. The other components, such as avionics, mostly come from the 747's as well. Even the landing gear are just the 747 gears. Literally the only parts that are new are the airframe-which are built of carbon fiber to save weight. However with 2 extra engines they would have wider performance margins which would mean there'd be less need for weight-savings in the first place. Even a wing made of lead will fly if you move it fast enough- and if they're really not moving fast enough that they need to seeep the wings with 6 engines (which I find a doubtful explanation) then 8 engines certainly won't push them supersonic... Another possible explanation I thought of (why am I the only one pushing actual reasonable explanations here? I came here looking for ideas better than my own- or better yet somebody who actually KNOWS the reason and has sources...) is that perhaps they kept the wings straight because they want to make the aircraft easier to fit in a hanger, and bring out of it... The suborbital plane? YNM, the Stratolaunch 351 is a plane designed for releasing *ORBITAL ROCKETS* that carry satellites all the way to Low Earth Orbit, not smaller suborbital planes. They haul the rocket up above the densest part of the atmosphere so it doesn't have to climb as far and its engines become more powerful/efficient due to lower ambient pressure. That, and the ability to fly to different lattitudes before release are literally the main reasons for air-launch in the first place- so the ability to fly higher (due to increased speed) would be an undeniable advantage.

No Derek, it's 100% clear exactly what we were discussing. I said the engine ate designed to operate up to at least Mach 0.86, citing that they fly this fast on the 747, and you said "Or that the top speed is considerably slower than you've presumed it to be." You were wrong, the engines clearly DO operate up to those speeds, and you should admit it. I strongly dislike your insistence that you are never wrong about anything, and if you really must follow me around posting on all my threads (I can't stop you- though you've been doing it for months now) then you really ought to admit you're wrong now and then... As for the speed the aircraft actually flies at, we literally know nothing about it. But swept wings are an advantage in pretty much any design where runway clearance is ample and getting shot at is not a concern (so you don't need to worry about how overall shape affects how easy your plane is to shoot), as it provides a further-back Center of Lift and more rear edge space for trailing edge control surfaces to attach to with literally no penalty to drag (which is determined by aspect ratio and frontal area- both of which are unaffected by wing-sweep as long as you keep wing-area and span constant). I'm not talking large delta wings with a high chord here- I'm talking relatively thin wing tips with a low chord. The only reason I can think of not to do this is that the bending moment is too high this way (swept wings increase how much torque is placed on wings as they are further from the attachment point at the ends...) Then again, maybe I just answered my own question. Alternatively, since Scaled Composites is building most of the airframe out of identical carbon fiber panels, maybe adding wing-sweep would have required a new type of panel, driving up costs. Think of needing a new wing part type in KSP to build a desired shape- if you don't already have it unlocked in Career Mode, getting it is going to cost you money...

If it was about stability then sweeping the wingtips would be even MORE important, since it moves the Center of Lift back. There has to be more to it than that...

The *engine* top speed Derek. That is clearly at least Mach 0.86 as a 747 flies at Mach 0.86, regardless of the capabilities if the airframe. You're wrong in that sense, and I'd like you to admit it. The point of flying faster, on the other hand, would be to fly higher. If you fly faster with only a small increase in drag and mass due to an extra pair of engines, you can fly higher for the same wing-area. Flying higher and faster not only means reduced Delta-V to orbit, it means higher thrust and USP for the rocket engines due to reduced ambient pressure. Every little bit helps with spaceflight, and even a 1% increase in payload capacity relative to what it was before (say 25250 kg instead of 25000 kg) can be worth hundreds of thousands of dollars... If they need as much power on the runway as possible (which I don't doubt- it IS a massive aircraft) then why use only 6 of the engines instead of all 8? Altitude isn't such a big problem on the other hand. They intend to release the payload at 30,000 feet- which is LESS than the roughly 45,000 foot altitude-ceiling of the 747's they were cannibalizing. Plus, as I've stated *repeatedly*, the faster you fly the higher you can fly with the same wings...

The engines are designed to perform well at Mach 0.86 and they have access to up to 8 of them from the two 747's they cannabilized- yet only choose to make use of 6 of them on the Roc. If it was a matter of needing more thrust to fly faster, they could just strap on a couple more of the engines. That they don't indicates they provably already have enough thrust to fly at the engines' designed top speed...