Sideways Engineering the SpaceX Dragon

[gunsandrockets]

This...<br /><br />http://sbir.gsfc.nasa.gov/SBIR/abstracts/00/sbir/phase1/SBIR-00-1-20.02-8728.html<br /><br />...suggests an ISP over 300s.<br /><br />This link...<br /><br />http://66.102.7.104/search?q=cache:BSnI_9mj4a0J:www.mesasphere.com/Mark.htm+lockheed+nitrous+oxide+monopropellant&hl=en&gl=us&ct=clnk&cd=2&ie=UTF-8<br /><br />...suggest an ISP of 320s.

 

[mrmorris]

<font color="yellow">"Though I suspect NASA would insist on using their own launch facilities for ISS missions..."</font><br /><br />Anything is possible. I think they'd be more interested in getting the maximum launch windows possible. After all -- they have to fly the astonauts out to KSC a few days prior to launch. Why should flying them out to Kwaj be any different -- seeing as the hardware isn't theirs anyway.<br /><br />I simply can't see the Dragon lifting a crew of seven *and* having any waste-handling capability. If it doesn't have that, then I can't see missions over about six-eight hours from launch to docking being feasible. If missions over that length of time aren't feasible, then there has to be some means of getting the maximum number of launch windows possible so that filtering out all but the shortest becomes reasonable.<br /><br />I didn't pick a flight time restriction that short because I think it'd be 'neato' for the crew-exchange vehicle to get to the ISS faster than the STS or Soyuz. I picked it because I don't see an alternative once you walk donw the path of what's possible given the volume available in Dragon.

 

[mrmorris]

OK -- we're pogoing each other's posts. <br />So does Zubrin's N20-Hydrocarbon work or not? <img src="/images/icons/smile.gif" />

 

[gunsandrockets]

"I didn't pick a flight time restriction that short because I think it'd be 'neato' for the crew-exchange vehicle to get to the ISS faster than the STS or Soyuz. I picked it because I don't see an alternative once you walk donw the path of what's possible given the volume available in Dragon."<br /><br />Depends on the acceptable level of comfort. Astronauts spent up to two weeks in the Gemini without a zero G toilet (that must have been interesting). If a particular Dragon flight is packed with seven people and they have to spend two days to reach the ISS, they will have to make do just as the Gemini crews did. Though I don't see why Dragon missions carrying a more ordinary complement of 3 people couldn't be flown. Maybe the simplest solution is to pack the Dragon with 7 people only for quickie flight launch windows and fly 3 during all other launch windows.

 

[gunsandrockets]

"So does Zubrin's N20-Hydrocarbon work or not?"<br /><br />Well if you can believe the information provided by the ex Pioneer Astronautics engineer, the Zubrin mixed monopropellant isn't very promising after all. Shuttle Guy's skepticism was correct.<br /><br />But the ISP that monoprop mixture provided seems relevant to a bi-prop N2O + hydrocarbon rocket, an ISP of 320s.

 

[mrmorris]

<font color="yellow">"Though I don't see why Dragon missions carrying a more ordinary complement of 3 people couldn't be flown. Maybe the simplest solution is to pack the Dragon with 7 people only for quickie flight launch windows and fly 3 during all other launch windows. "</font><br /><br />I agree that with less people, waste facilities become more feasible. I beleive I made a brief mention of that possibility earlier in the thread. However, there's a problem with the assumption of three people being the 'ordinary complement'. Unless the ISS construction stalls, by the time NASA would be looking at the Dragon for crew rotation, the crew in question won't be three people anymore. The remaining labs should be in place, along with the US oxygen generator & water recycling system. Everything is *supposed* to be in place for a crew of six-seven. While it would be possible to rotate the crews by halves -- the Dragon-as-lifeboat requires that it be able to evacuate the station.<br /><br />Obviously the ISS completion has not happened, and may <b>not</b> happen, but in designing Dragon, SpaceX has to make the assumption that it will. Assuming that three people and a toilet will be acceptable simply won't fly.

 

[gunsandrockets]

a few points:<br /><br />1) The COTS contract for SpaceX is to provide ISS cargo. Providing crew is something down the road.<br /><br />2)ESAS documentation for the block 1 production CEV for crew access to ISS defined the crew as three people, suggesting normal ISS crew rotations of only 1/2 and not the entire crew of the ISS at once. This makes sense as the ISS is just that, International. Which means Russia as well as NASA will deliver crew to the ISS when it is complete. It also suggests a Russian and an American spacecraft will be docked to the ISS at the same time, splitting lifeboat functions.<br /><br />3)The seven man crew design of the Dragon was before SpaceX won the COTS contract. Though no doubt SpaceX hoped for NASA money when they first designed the Dragon, the seven crew specifications probably has much more to do with SpaceX's cooperation with Bigelow Aerospace. So the seven crew function was probably designed to accomodate Bigelow's needs rather than NASA's.<br /><br />"Assuming that three people and a toilet will be acceptable simply won't fly."<br /><br />Even if true, then the seven people will have to settle for crapping their diapers. Hardly a deal breaker. <br /><br />

 

[mrmorris]

<font color="yellow">"Providing crew is something down the road."</font><br /><br />SpaceX <b>wants</b> that crew contract. To get it, they still need to make sure that the Dragon capsule design takes crew requirements into account from the outset. Failing to do so will generate some big expenses down the road if they have to re-engineer parts of the craft and then re-tool to support the new designs.<br /><br /><font color="yellow">"Even if true..."</font><br /><br />We'll have to agree to disagree. This is especially true as I'm not really sure anymore what we're disagreeing about. You've accepted that there are launch windows where the timeframe for getting Dragon to the ISS is fairly quick (sub 6-to-8 hours). Crew rotations happen every six months and aren't generally time-critical (i.e. going to seven months wouldn't be a disaster). Presumably they'll be able to find enough windows each year to support that schedule.

 

[mrmorris]

<font color="yellow">"One other obstacle to available launch windows is the Russian launch schedule of Soyuz and Progress spacecraft to the ISS. To that must be added the future ESA launch of the ATV spacecraft and JAXA launch of the HTV spacecraft to the ISS. I know NASA likes to keep the traffic clear of simultaneous launches to the ISS. "</font><br /><br />I meant to respond to this separately last night. We got so caught up in our pogoing about the N20-Hydrocarbon issue that I forgot. <br /><br />There are clearance issues with the orbiter being docked and the use of one of the three russian docking ports. Shuttle guy has spoken of the exact issue on more than one occasion, but I forget the specifics. NASA obviously would not like to have two spacecraft in free-flight near the ISS at once, but I can't see an issue once one has docked/berthed.<br /><br />The current design of the Dragon has it berthing at the CBM -- where the only user is the MPLM (which will be no more once the shuttle goes away). JAXA's HTV is supposed to use it, and might be by the time Dragon exists, only time will tell. Progress and Soyuz should be non-issues as the docking ports are nowhere close to the CBM.<br /><br />Beyond that -- Elon has indicated that the Dragon was designed to fit the CBM with the express intention of making it large enough for <b>any</b> of the attachment mechanisms. It could be fitted with either a russian P&D or with the APAS-89. Fitted with the APAS -- it should have essentially no competitors for slots, as the orbiter is the only user for that style. The only other *possible* user that I could envision would be the CEV, and if the Dragon will be covering crew rotation -- why would NASA pay to send a CEV up there? Given the <b>incredibly</b> smaller footprint of the Dragon when compared to the shuttle -- the clearance issues that exist between it and the third russian port should be a non-issue.<br /><br />Of course if SpaceX end up getting ports o

 

[nyarlathotep]

gunsandrockets: "If a particular Dragon flight is packed with seven people and they have to spend two days to reach the ISS, they will have to make do just as the Gemini crews did."<br /><br />Not at all. If they managed to squeeze this into a soyuz capsule, I'm sure they could bolt a similar contraption into dragon.<br /><br />Morris: "Of course if SpaceX end up getting ports of all three styles, and either has Dragon capsules with each, or even better -- makes them modular such that they can be switched out, then their operational flexibility becomes immense."<br /><br />I have strong doubts that the Russians would let them dock on their ports.

 

[mrmorris]

<font color="yellow">"I have strong doubts that the Russians would let them dock on their ports. "</font><br /><br />I don't know what the international agreements are on that point. That ATV, of course, will be docking at a russian port. I don't really know why you think it would be a problem. If the agreements allow, I would not be suprised at all if the russians were to try to charge NASA a docking fee to use their ports. I suspect, however, that this would not go over well with NASA. This would be especially true since the Zarya module contains (IIRC) two of the three russian ports and the Zarya module was funded by NASA.<br /><br />In any event, the russian docking ports would actually be the least attractive to my mind. While there's three of them, the ATV, Progress and Soyuz will all be competing for space on them. My preferred second choice after the CBM would be an APAS port on the Dragon. Unless something is built to use the APAS -- it will become a wallflower.

 

[mrmorris]

While I was doing some research on the CBM to answer the above question -- I found the following interesting paragraph and graphic on the JAXA site. It's amazing what Google will get you when you're not even looking:<br /><br /><i>"The CBM is a commonly used mechanism to connect non-Russian pressurized modules. Two pressurized modules will be brought to the connection position using a robot arm, and the each modules latch mechanisms will be connected by driving a motor which includes active CBM(ACBM). Once the connection is established, the CBM will Provide the resource connection such as, electrical power, communication signals, and fluid between the connected modules. It also serves as a path for the crew. A total of six CBMs will be mounted on each nodes. At present, only Unity (node-1) has been launched..."</i><br /><br />The part about electrical power was interesting to me. I'd speculated earlier in the thread about the need for a 'power cord' between the destiny laboratory and the Dragon. Except... if the CBM port already provides a power connection, then the Dragon has access to ISS power immediately after berthing. This further baffles me why the "mouse ear" solar panels are even in the design.

 

[mrmorris]

As continues to happen -- while immersed in the Internet looking for one thing, I found another. I can't imagine how I missed it before, seeing as it's a Gemini-factoid and I really thought I was fairly well studied up on that particular craft. However, apparently it never registered with me that the Gemini OAMS ran on LOX/Kerosene. Specifically, it used the Rocketdyne SE-7 Series, like the SE-7-100 (There were several thrusters of different sizes in the OAMS).<br /><br />So... LOX/Kerosene. When thinking about non-hypergolic combinations, I've always focused on LOX/Methane... because NASA is focusing on that combination. However, they're interested in it basically because they want in-situ propellant generation capabilties for Mars. That's not a a factor for Dragon, <b>and</b> SpaceX specializes in LOX/Kerosene thrusters. Given an existing thruster to reverse-engineer; their work to date designing two LOX/Kerosene thrusters; the logistical advantage of <b>everything</b> running on the same propellant/oxidizer combination; and their goal of producing as much of the spacecraft in-house as possible to minimize costs -- I have to consider it most likely that they'll go this route. <br /><br />The question remains, though... what will they name it? The Kestrel is the smallest falcon in the US. Googling for the smallest bird of prey, I came up with the Black-Legged Falconet (Micrphierax fringlius). So will we be seeing information at some point soon about SpaceX's new Falconet RCS thruster? <img src="/images/icons/smile.gif" />

 

[soyuztma]

You could also use hydrogen peroxide/kerosene for your thrusters. It's not toxic and not cryogenic, so no storage problems. But i didn't directly found an example of an peroxide/kerosene RCS. So i have no idea if you can buy such a system of the shelf.<br />But i don't think they should build their own RCS system, there are enough companies that should be able to build one for a decent price. <div class="Discussion_UserSignature"> </div>

 

[gunsandrockets]

"We'll have to agree to disagree."<br /><br />I'll agree to that.<br /><br />"I'm not really sure anymore what we're disagreeing about."<br /><br />You think the Dragon will carry seven crew to ISS and therefore inadequate room for toilet use. You think that mandates a Dragon flight time of mere hours. You think available launch windows to ISS makes this easy.<br /><br />I admit that the launch window issue isn't a roadblock as I first thought, though I still think it's harder to schedule than you do and more importantly it's an unneccesary nuisance to wrap the Dragon flight schedule around that flight time. Because I think crew size and/or toilet availability will not drive the mission design. <br /><br />

 

[gunsandrockets]

"So... LOX/Kerosene... SpaceX specializes in LOX/Kerosene thrusters. Given an existing thruster to reverse-engineer; their work to date designing two LOX/Kerosene thrusters; the logistical advantage of everything running on the same propellant/oxidizer combination; and their goal of producing as much of the spacecraft in-house as possible to minimize costs -- I have to consider it most likely that they'll go this route."<br /><br />Very reasonable. The LOX storage could perform double duty for air supply. <br />

 

[gunsandrockets]

"Not at all. If they managed to squeeze this into a soyuz capsule, I'm sure they could bolt a similar contraption into dragon."<br /><br />Squeeze? The Soyuz toilet was placed in the Orbital Module, not the Descent Module. And actually the Soyuz is quite a roomy spacecraft compared to the Apollo Crew Module. It's only the Descent Module by itself that's cramped when holding all three crew.<br /><br />Plus the real issue for using a toilet in a spacecraft isn't the size of the hardware needed as much as it is the space for a crewmember to maneuver around inside cramped quarters. The Dragon capsule holding seven crew is packed pretty tightly.

 

[mrmorris]

<font color="yellow">"But i don't think they should build their own RCS system, there are enough companies that should be able to build one for a decent price. "</font><br /><br />If you can find me one for 'a decent price', I'd be very interested. I've done a lot of searching and haven't found anything but hypergolic reaction control systems. Before this afternoon, I was convinced that SpaceX was going to have to go COTS hypergolic -- at least for the first version of Dragon. Even then, there's nothing like a 'standard' COTS RCS that SpaceX can pick out of a catalog. The RCS has to be custom built for Dragon... which means expensive. Most COTS RCS systems are for satellites -- they are for adjusting attitude rather than changing orbits. This means the thrusters are too low powered. A company will be glad to make larger thrusters... for a price.<br /><br />While I was researching RCS engines today, I pulled up the SE-7-100 and realized this was Gemini's OAMS engine. Bright lights and fireworks went off as it all came together. It's going to take a really good argument to convince me this isn't the route SpaceX will go for. If something absolutely prevents them from having an in-house LOX-Kerosene RCS ready for V1 of Dragon, then V2 will have it. The logistics advantages of a non-toxic combination that matches SpaceXs engineering capabilities and their existing oxidizer/propellant infrastruture is simply too incredible to assume they won't go this route. The added expense of developing the system in-house will be made up rapidly by not having to duplicate propellant storage systems, procurement contracts, etc. -- not to mention that once they've developed the system, they'll be able to build subsequent systems for a fraction of what someone would charge them. Assuming SpaceX is successful in this, there's not going to be one Dragon capsule or two or three -- there's going to be a half-dozen or more. If they hadn't already built the Kestrel and t

 

[gunsandrockets]

"NASA obviously would not like to have two spacecraft in free-flight near the ISS at once, but I can't see an issue once one has docked/berthed."<br /><br />Available berthing ports was not the factor I thought could interfere with Dragon launches.<br /><br />It's true that on average, 25% of available launch windows from Florida can provide short flight times to the ISS. And if only the Dragon were flying to the ISS it would be ridiculously easy to schedule a launch.<br /><br />But many other spacecraft will fly to the completed ISS besides manned Dragon flights. There will presumably be unmanned Dragon cargo flights to ISS, probably many more than manned flights. Add to that the launch of all the other international partner's spacecraft, manned and unmanned, to the ISS.<br /><br />Every one of those flights will eat up available launch windows for the manned Dragon, because each of those flights will probably take one or two days to reach the ISS as is current standard practice. And since short launch windows to the ISS are not spaced evenly in time but are clustered instead, finding the short flight time launch windows for manned Dragon flights will take some effort. It won't be easy.<br /><br />For example, let's say one particular ten day period has multiple short flight time launch windows for a scheduled Dragon flight. But then there is some trouble with a prior scheduled Soyuz or Progress mission which delays the flight for a few days. Then there goes the short flight time window for the Dragon.<br /><br />The point is traffic to the completed ISS could be so heavy that the available launch windows for short flight times to ISS will be limited. Even so it should be easy to find at least two windows per year that match short flight times.

 

[mrmorris]

<font color="yellow">"The point is traffic to the completed ISS could be so heavy that the available launch windows for short flight times to ISS will be limited."</font><br /><br />Let's look at the traffic patterns: Currently there are the following flights to the ISS (excluding shuttle flights)<br /><br />3-4 Progress flights per year (we'll call it 4)<br />2 Soyuz Flights per year<br /><br />Presumably in the future, there will be the ATV and HTV making regular flights. I know the ATV is supposed to have a six-month orbital lifetime, so we'll assume there are two per year. I believe this is considerably overstating the case, since they're expendable and I think only three or four are actually planned to be built -- but we'll assume two per year for it, and another two HTVs annually.<br /><br />2 ATV flights per year<br />2 HTV flights per year<br /><br />I don't really buy the concept of Dragon cargo becoming a launch constraint on Dragon crew because it won't be in competition for the same set of launch windows (it can take several days) and I have difficulty envisioning a failure that would cause a conflict under anything but truly exceptional conditions. It's possible to get struck by lightning... but worrying about it isn't really practical.<br /><br />So we have ten flights a year going to the ISS in this future world that can cause conflict. None of them are launching from KSC or Kwaj, so they're not going to the <b>same</b> launch windows as Dragon. Each takes ~2-3 days getting to the ISS, for a grand total of 20-30 days out of the year -- less than 10% at the highest. I don't see this being a particular problem cancelling out launch windows. <br /><br />As to a Soyuz/Progress delay causing a problem -- the 6-8 hour launch-to-berthing time provides another advantage. Unless that delay is Twilight Zone freaky accurate -- it's not going to cause a problem. It has to be delayed in such a fashion that the two will be in close proximity to the ISS within

 

[mrmorris]

The following is the current Equipment List I have in the application for Dragon. It fluctuates constantly as I add new things or tweak existing numbers. Right now I'm working to remove the 'RCS' line item. Essentially it's just a placeholder with the mass taken from the Apollo CM RCS figure, the volume a WAG, and the power assuming 45-watt thruster valves with a maximum of four firing simultaneously. I've found a couple of documents on RCS design and mass estimation, so I'm going to try to come up with a figure that's actually based on a broader set of data.<br />

 

[mrmorris]

Well Crud! Apparently Astronautix is wrong. Looking at the Gemini Familiarization documentation, the thrusters ran on N2O4/MMH. *sigh*<br /><br />I still **really** like the idea of the LOX/Kerosene RCS. However -- I no longer have the assurance that "It has been done -- therefore it can be done." Unless I can find some LOX/Kerosene thrusters in 25-pound and 100-pound classes, it's conceivable that it's simply not an option.

 

[mrmorris]

Well -- looking for RCS-sized LOX/Kerosene thrusters, I found two besides the (mistaken) Gemini OAMS thrusters. If I assume <b>these</b> have the correct propellant listed, then it's at least possible to use LOX/Kerosene for the required class of thruster. The relative scarcity of LOX/Kerosene thrusters in this force class makes me wonder wabout the downsides.<br /><br /><i>Manufacturer Name: RS-34. Designer: Rocketdyne. Developed in: 1978. Propellants: Lox/Kerosene. Thrust(vac): 304 N (68 lbf). Isp: 255 sec. <br />Peacekeeper Postboost Attitude Control. Pressure-fed. 8 attitude thrusters in each postboost propulsion system. <br /><br />Manufacturer Name: P4-1. Government Designation: LR-64. Designer: Rocketdyne. Developed in: 1963. Application: AQM-37. Propellants: Lox/Kerosene. Thrust(vac): 470 N (100 lbf). Isp: 262 sec. </i><br /><br /><br />I may be whistling in the dark, but I'm going to continue to assume this is a possibility and press on with the design of a SpaceX LOX/Kerosene RCS.<br /><br />Looking at the Kestrel -- it's a reasonably good model for what SpaceX needs in an RCS. Mind you, it's on a scale 70 times too large. <img src="/images/icons/smile.gif" /> Kestrel is ablatively cooled in the chamber and radiatively cooled in the nozzle. I would assume that the Falconets will be completely radiatively cooled. The Kestrel's vector is controlled by electromechanical actuators which obviously won't be used in an RCS. For ignition, the Kestrel uses LOX/Kerosene torch igniters to allow for as many restarts as required (obviously a critical issue for an RCS/OMS).<br /><br />In designing RCS thrusters for Dragon, I expect that SpaceX would create two two thrusters in ~25 and ~100 pound classes (the Falconet-25 and the Falconet-100). Each of the four RCS clusters on the Dragon would have one Falconet-100 facing aft, and three Falconet-25 thrusters facing forward and to the sides.<br /><br />To generate WAG specifications for the Falconet thrusters, I'm going to assume

 

[mrmorris]

I contacted a engineer that specializes in propulsion systems and presumed to ask him about the possibility of RCS-sized LOX/Kerosene thrusters. He replied indicating that there is no engineering reason that thrusters of that class cannot be made using LOX/Kerosene. He beleives that the main problem will revolve around zero-G and getting LOX fed to the thrusters because it cannot be pressurized in the same fashion as a non-cryogenic oxidizer. In addition, he indicated that the LOX in the feed lines running to the thrusters will gassify, meaning that for short bursts especially, what will get to the thrusters isn't LOX but GOX.<br /><br />I have to wonder if necessity can't be turned into a virtue. Instead of making LOX/Kerosene thrusters, SpaceX can make thrusters designed for GOX/Kerosene. Feed the LOX to an intermediary tank to gassify and pressurize it -- then use <b>that</b> tank to feed the engines. Dragon needed a GOX tank anyway to handle the LOX boiloff and to use as an oxygen replenishment source for the ECLSS. That tank could also be used as a heat-sink by the ECLSS to help dump excess heat from the Dragon cabin.

 

[mrmorris]

Comparing all of the graphics that SpaceX has released, one of two things needs to be shot: either my concept of where they're putting the parachutes, or the person who made the graphic. Based on the semi-transparent views of Dragon -- the location where I thought parachutes might be placed does not exist.<br /><br />Because the exterior views are all in 2-D and the transparent views all in a 3D and angled, I hadn't noticed until this morning that the 'dome' doesn't appear to hinge where I thought. From the graphic below -- I've always assumed a hinge at 'Location 1', whereas the transparent view shows it at 'Location 2'. Given a hinge at location 1 -- I've assumed that the CBM was at location 'A', whereas apparently it's at 'Location B'.<br /><br />The upshot is that there's more volume in the dome than I expected and less in what I was calling the forward equipment section. What's unfortunate is that I can't really see much use for the additional volume in the dome. The fact that it must open and close makes it impractical to put the parachutes there -- or most equipment for that matter. I'll have to give it some thought. I have low hopes for deducing the method of their madness, however, since I still can't even figure out what they were thinking about with the mouse ears.