Sideways Engineering the SpaceX Dragon

[mrmorris]

There's an article on Hobbyspace about Almaz Excalibur -- a company marketing themselves as a space tourism company with plans to use the old Soviet Almaz station concept. This has no relation to the Dragon, of course, but it's always instructive to do comparative analysis with other capsule designs. Their Powerpoint here indicates they are 'hardware rich' with eight capsules and a space station. I have to take that with a pretty big grain of salt, since the hardware is all from the mid-60's and the Soviets never even flew the capsules manned. I also can't **imagine** they could be flown without massive refits of the electronics, avionics, and other subsystems -- especially as most likely never had a full set in the first place. However, I was interested enough to do some research on the Almaz capsule -- namely the TKS VA. The stats listed in the PPT match the Astronautix specs exactly.<br /><br /><font color="orange">TKS VA Capsule</font><br /><br /><b>Crew: 3<br />Launch Mass: 7,300 kg<br />Mass in orbit: 4,250 kg<br />Landing Mass: 3,800 kg<br />Total internal Volume: 8.37 m3<br />Habitable Volume: 4.56 m3<br />Length: 3.64m<br />Diameter: 2.79m<br /><br />Crew mass: 255 kg <br />Miscellaneous Contingency: 50 kg<br />Payload: 1,880 kg<br />L/D Hypersonic: 0.25</b><br /><br />The Astronautix article mentions an interesting bit about the landing method:<br /><br /><i>"Once the capsule was subsonic a drogue parachute deployed for seven seconds, followed by the main chute with 1770 square meters of area. The capsule made a soft landing using a retrorocket in the parachute lines. This was triggered by the Probki radioactive sensor system within the Kaktus gamma ray altimeter, which set off the DU braking unit for a soft landing of the capsule."</i><br /><br />I must admit I wou

 

[mrmorris]

The design elements of the TKS VA capsule have been rattling around in my head the past couple of days. I've decided that I like the concept of a propulsion section that is completely separate from the crew capsule. I can't see going with the forward Propulsion/Equipment section like the TKS VA, but if the rear propulsion/equipment section of Dragon were modularized to the point of being removable by a double-handful of bolts, I can see some interesting possibilities.<br /><br />For one -- in the event of a low-altitude abort, explosive bolts could detach the crew section from the propulsion section. A quick calculation of the affected elements indicates this would drop the mass to be accelerated in an abort by fifty percent or more. Since low-altitude aborts are what drives the size of the launch-escape motor, this would cut the required propellant load massively. There's added complexity involved, more failure modes, and the rear-bulkhead of the crew cabin would have to be beefier to handle the landing impact... but it's an interesting thought.<br /><br />To me, however, the more interesting possibilities come from the operational advantages this would provide. Aborts, after all, are an anomoly. You have to plan for the possibility, but the *probability* is that they will happen very seldom if ever. Operations, on the other hand, happen every flight... by definition. So what exactly does a propulsion module (PM) that's designed to be modular buy in terms of operations?<br /><br />- <b>Enhanced reusability</b>. The portions of the craft that will be beat up the *most* on every flight is the PM. The thrusters undergo wear, the TPS wear is obvious (I hope), this section takes the landing impact, etc. By contrast, the forward portions of the capsule undergo much less wear and tear. Building in the ability to replace a tapped-out PM on a given capsule easily will seriously increase the reusibility of the remaining elements.<br /><br />- <b>Increased redundancy</b>

 

[erikm]

<blockquote><font class="small">In reply to:</font><hr /><p><i>"Once the capsule was subsonic a drogue parachute deployed for seven seconds, followed by the main chute with 1770 square meters of area. The capsule made a soft landing using a retrorocket in the parachute lines. This was triggered by the Probki radioactive sensor system within the Kaktus gamma ray altimeter, which set off the DU braking unit for a soft landing of the capsule." </i><br /><br />I must admit I would never have thought to add a retrorocket **in** the parachute lines to further reduce landing speed. Mind you -- I'd have to actually view the engineering of it before I'll be able to consider it without shuddering at the thought of what an aiming malfunction in said method would do to the parachute...<p><hr /></p></p></blockquote><br />It reads a lot like the retro system the russians use to airdrop BMD armored personnel carriers. Basically it looks like a rocket pack mounted at the top of the parachute risers, below the suspension lines. It goes off shortly before landing to soften the impact. Apparently the landing is soft enough that troops can ride inside the BMD instead of having to jump separately (which is the norm). <br /><br />Adapting the principle to spacecraft isn't that much of a leap. I don't know anything about the technical details though.<br /><br />Cheers,<br /> ErikM<br />

 

[mrmorris]

The article on SDC Space News today about TacSat2 and their Inertial Stellar Compass was interesting.<br /><br /><i>"If tests planned for the weeks ahead verify the Inertial Stellar Compass’s performance under a variety of mission conditions, spacecraft designers will have at their disposal a flight-verified attitude determination system that offers significant mass and power savings. Fuhrman said the 2.9 kilogram instrument package runs on 3.6 watts of power — less than half the mass and power of conventional systems."</i><br /><br />I doubt SpaceX would want an experimental tech to be their primary attitude determination system, but it would make a highly qualified secondary or tertiary. Also of interest is the following:<br /><br /><i>"NASA struggled for several years to find flight accommodations for the Inertial Stellar Compass. The space agency originally planned to fly the instrument on board the space shuttle, but it was bumped from the manifest even before the 2003 Space Shuttle Columbia accident made it even more difficult for non-space station payloads to find a berth on the shuttle. Fuhrman said NASA looked into flying the instrument on a proposed commercial solar sail mission and one of its own Earth science satellites before finding the TacSat opportunity."</i><br /><br />If SpaceX were to design the Dragon with the express intent that tertiary systems are <b>intended</b> to be swapped out on a mission-by-mission basis -- this would be an excellent means of adding to the commecial prospects of Dragon. It would allow them to charge NASA for acting as a testbed for new equipment such as this and it would give them access to the most advanced hardware being developed for spacecraft.

 

[mrmorris]

I've been too busy with actual real-life recently to do justice to a Dragon analysis post recently. For that matter, I am today as well. However, this article about the Orion CEV parachute testing is something I want to go over in relation to the parachute calcs I've done for Dragon. I wanted to get the post in here so it's convenient when I have time to do some comparative analysis.

 

[vulture2]

Intersting concepts. Two thoughts:<br /><br />Mir flew in solar-inertial mode (the vehicle itself oriented so that solar panels were at optimal angles to the sun) except during docking, so orientation of the solar panels for max power is possible most of the time even if they are not fully steerable.<br /><br />The Soyuz uses one chute with one reserve; this might be considered as an alternative to the three chute Apollo system; in theory two parachutes (out of three) would be smaller for the same size payload than a single chute, but this is offset somewhat by complexity and the angles at which multiple parachutes must fly.

 

[mrmorris]

<font color="yellow">"Mir flew in solar-inertial mode..."</font><br /><br />Anything is possible. The graphics from Spacex indicate that solar panels will be there. Even give a flightpath that maximizes solar flux on the panels, I simply can't see anything that justifies their mass. Given the power density of Li-Ion batteries, there's plenty of juice for the flight *to* the station, and once *at* the station, using ISS power makes much more sense. Soyuz uses much more power than Dragon will; I doubt the power density of its battery system is up to modern standards; and I don't beleive the Probe and Drogue docking ports have the ability to access ISS power. If Dragon doesn't *need* the panels, then I can't see SpaceX spending the engineering time and money to add them. I may be wrong... I often am.<br /><br /><font color="yellow">"...this might be considered as an alternative to the three chute Apollo system..."</font><br /><br />Despite the above example, I really <b>am</b> trying to stay away from *designing* the Dragon. The one graphic I have from SpaceX of a Dragon splashdown shows a 3-chute Apolloesque parachute system. The Apollo methodology is well designed and robust, so I expect that's what they'll go with.

 

[bitbanger]

<blockquote><font class="small">In reply to:</font><hr /><p>If Dragon doesn't *need* the panels, then I can't see SpaceX spending the engineering time and money to add them.<p><hr /></p></p></blockquote><br />The assumption here is the the only destination is ISS. Dragon was in development well before the COTS win. It stands to reason that other uses are being considered/planned.<br />

 

[mrmorris]

<font color="yellow">"The assumption here is the the only destination is ISS."</font><br /><br />NO! I already responded at the beginning of this thread when you made essentially the same statement. The assumption here is that the <b>first</b> destination is the ISS. It's difficult enough trying to sideways-engineer the version one Dragon given the snippets of data available on the Web. It's <b>impossible</b> to sideways-engineer a later version of said craft when much of the engineering involved in creating *it* will stem from problems/enhancements that become obvious due to the flight data from V1.

 

[soyuztma]

Hobbyspace has posted a link to the Space Act agreement between NASA and Rocketplane Kistler and NASA and SpaceX. Page 84 and 85 gives some technical details about the Dragon: <br />-"The primary Dragon propulsion system consists of four independent modules containing four 50 lbf NTO/MMH thrusters and four 2 lbf cold gas (He) thrusters, which achieves thruster out redundancy."<br />-"Total NTO/MMH mass is 2000 lbs, sufficient for on orbit maneuvering, deorbit and re-entry control."<br />-"SpaceX possesses deep internal expertise in developing space qualified NTO/MMH thrusters" This probably means they are going to build there thrusters themselves.<br />-"The spacecraft is designed to have a canonical blunt body, lifting re-entry with a radially offset center-of-mass, which induces a trim angle of attack. The sidewall angle of 15° is a compromise between intemal volume, TPS mass and L/D. The L/D of approximately 0.3 allows a 2° re-entry corridor and keeps nominal deceleration for the crew safely below 5g"<br />These numbers are probably allready seriously out of date, but they probably give a good indication.<br /><br />The document also includes the different milestones and the money they get when they achieve a certain milestone. <div class="Discussion_UserSignature"> </div>

 

[mrmorris]

Cool. I've got to get back into this soon and do some more analysis. There was one bit that you missed about the thrusters ... which makes it even more interesting:<br /><br /><i>"In addition, there is an auxiliary set of two NTO/MMH thrusters for two fault tolerance near the ISS."</i><br /><br />So there's the four primary NTO/MMH quads, plus four cold-gas (quads?), plus two more NTO/MMH thrusters. I'll have to pore over the graphic and see if any of the backups are apparent.<br /><br /><i>"Crewed launches are in addition to this, nominally 2 or 3 per year. A launch with a crew of 3 can also include up to 1,000 kg of cargo (in addition to the 250 kg pressurized ISS cargo mandated in the COTS SRM Summary), which can be divided between pressurized and unpressurized as needed. A launch with a crew of 7 will have essentially no additional cargo capacity."</i><br /><br />These should help firm up some of my mass guesstimates. If four people are displacing 1000kg of cargo, then I have a reasonable read on the mass of their life-support requirements. Normally astronaut mass is assumed to be in the neighborhood 200lbs/90kg. Four people then mass at about 400kg, so something is using up another 600kg.<br /><br /><i>"The primary pressure vessel will be made of welded aluminum and the deployable nose cone of lightweight carbon composite materials. Deployables have been kept to a minimum for reliability. The only external deployment mechanism is the hinge and latch mechanism for the nose cone, with solar arrays hidden inside. If the closing and latching procedure does not complete nominally, the entire assembly can be severed at the hinges by redundant pyro-cutters, since the nose is not required for safe reentry.<br />...<br />A CBM berthing interface is mounted under the nose cone which hinges open once in orbit."</i><br /><br />The problem of the nose cone not opening or closing was one I addressed at the beginning of the thread. Interesting that it's not required for re-entry. A

 

[annodomini2]

1 lb = 454gramms<br /><br />therefore 200lbs = 90kg. <div class="Discussion_UserSignature"> </div>

 

[mrmorris]

<font color="yellow">"therefore 200lbs = 90kg."</font><br /><br />Ayup... and I was in a hurry last night and converted <b>100</b> lbs to kg instead of 200 lbs like I meant to. Oops.

 

[mrmorris]

SpaceX just massively updated their website and they now have a page for Dragon that contains some more details. The only new engineering tidbit I see from the page is a high-level description of the Dragon structural elements.<br /><br /><i>"The Dragon capsule is comprised of 3 main elements: the Nosecone, which protects the vessel and the docking adaptor during ascent; the Pressure Section, which houses the crew and/or pressurized cargo; and the Service Section, which contains avionics, the RCS system, and other support infrasturcture."</i><br /><br />Durn it all -- they still haven't pinned down where the parachutes are getting placed.

 

[j05h]

Check out this pic of Dragon parachuting down:<br /><br />http://spacex.com/assets/img/Dragon-5.jpg<br /><br /><br /> <div class="Discussion_UserSignature"> <div align="center"><em>We need a first generation of pioneers.</em><br /></div> </div>

 

[mrmorris]

<font color="yellow">"Check out this pic of Dragon parachuting down"</font><br /><br />Ayup. That would be the same one I posted to this thread and analyzed last November. <img src="/images/icons/smile.gif" />

 

[soyuztma]

For some reason they have now placed the payload capacity number of the Falcon's in the FAQ section. It says that the Falcon 9 will be capable of putting 10350 kg in LEO. And this is about 1000 kg more than the last numbers they released. <div class="Discussion_UserSignature"> </div>

 

[spacester]

If ever there was a thread worthy of a bump just for the sake of being bumped, it would be this one.<br /><br />bump<br /><br />mrmorris, you only need 45 more posts to become a solar system! Surely there can be no higher priority in your life than that goal! <img src="/images/icons/tongue.gif" /><br /> <div class="Discussion_UserSignature"> </div>

 

[mrmorris]

*sigh* --- I've been ***really*** busy at work (i.e. the job where they send me actual paychecks). The Dragon Specifications thread almost dragged me in a couple of times. But I'll respond to the bump with a couple of 'quickies' that likely will turn into a 'longie' before I'm through. Bringing a snippet of that thread over:<br /><br />"In addition a Trunk Section is included, which provides for the stowage of unpressurized cargo and will support the solar panels providing power to Dragon." <br /><br /><b>gunsandrockets</b> noted a couple of things about that change. Namely that it makes the capsule closer to the BAR multi-role capsule and that the solar panels are no longer reusable. What strikes *me* about it is that it cuts out several of the arguments I had against the mouse ears. In a very early post of this thread, I prophesized that Dragon V1 wouldn't have ears. Ta Da ! My arguments were that it didn't need the power, or the added failure modes to docking/undocking and re-entry. Well -- with the panels in the 'trunk', the docking and re-entry issues disappear off the board. Also -- since the trunk is disposable, and will essentially be rebuilt for every single mission, SpaceX needs only include solar panels for missions where it makes sense to put up with the extra mass to gain the extra power. I've noted (and still believe) that for ISS missions, they should simply have enough batteries to ensure ample ergs for worst-case abort scenarios going up, and suck power from the ISS while berthed. <br /><br /><br />Very interesting were the new graphics that SpaceX added to their website. There's considerably more detail than in the earlier ones, which answers one question and poses others. One of the more puzzling ones to me is the nosecap -- it's been elongated. I haven't taken the time to measure how much, but it's obviously taller than it used to be. Is this just an aerodynamics change, or is something being stored there that neeed more room? If someth

 

[soyuztma]

<blockquote><font class="small">In reply to:</font><hr /><p>In the corresponding location on the bottom-right of the Dragon is a hatch/connector/doodad of some sort. I dunno about this one. My guess would be that's the port where they feed in external power/air until second prior to launch.<p><hr /></p></p></blockquote><br />The Dragon will be using the space station arm to dock so i'm pretty sure this is the grapple point for the arm. <div class="Discussion_UserSignature"> </div>

 

[mrmorris]

<font color="yellow">"... i'm pretty sure this is the grapple point for the arm."</font><br /><br />Certainly possible. My money's against it, though. All you need for the arm is an external hardpoint/handle. The area I was pointing to has a reasonable amount of interior volume being consumed.

 

[larper]

Doesn't the grapple assembly provide the capability of umbilical connections? If so, all that extra stuff would be the power and data transfer electronics and hardware. <br /><br />However, I would bet on the ground umbilical connection hypothesis.<br /> <div class="Discussion_UserSignature"> <p><strong><font color="#ff0000">Vote </font><font color="#3366ff">Libertarian</font></strong></p> </div>

 

[mrmorris]

<font color="yellow">"Doesn't the grapple assembly provide the capability of umbilical connections?"</font><br /><br />I don't *think* so. The arm is primarily designed to be moving large objects from point A to Point B -- solar arrays, trusses, MPLMs, etc. It's not intended to be an attach point to the ISS for an extended period of time, so I can't imagine it being designed with the massively increased complexity that would be required in order to act as an umbilical.

 

[larper]

I didn't mean that the arm was acting as an umbilical, but that the grapple fixture would include data/power connections FOR THE ARM ITSELF. One failure scenario is that once grappled to the capsule, the arm fails to be commanded to release from the station side controls. In this case, the astronauts inside the capsule would have to have the capability to command the arm from their side. This requires data/power connections to the arm.<br /><br />For a piece of payload, such a failure is not an issue and can be solved at leisure. This would not be the case for a manned capsule being manuevered by the arm. <br /><br />Just a thought. I still think it is a ground connection of some sort. <div class="Discussion_UserSignature"> <p><strong><font color="#ff0000">Vote </font><font color="#3366ff">Libertarian</font></strong></p> </div>

 

[mrmorris]

<font color="yellow">"I didn't mean that the arm was acting as an umbilical..."</font><br /><br />Ok... I'm not sure how I was supposed to understand that <i>"Doesn't the grapple assembly provide the capability of umbilical connections?" </i> didn't mean that you were expecting the arm to act as an umbilical. Notta problem though.<br /><br />At to control/power supply, IIRC the arm *does* have the capability to be controlled from either end. The ISS has several Power-Data Grapple Fixtures around the station, and the arm can 'inchworm' around from one to the next. Presumably, if the Dragon instead of having a simple handhold instead were equipped with a PDGF, it could conceivably send a command to the arm from on-board. Mind you -- there's no way it could actually *use* the arm, as the relatively low mass of the Dragon would mean that action-reaction would play merry Heck with the craft with any movement of the arm. <br /><br />However, I can't see much value-add. If the arm is ignoring signals from *one* PDGF, the chances that sending the same 'Let go of me' command from the other side would work seems remote. Unless the mass and volume of the PDGF in lieu of a 'dumb' grapple point is negligible, I don't see it.