Ben Brockert's Rocket Analysis

There have been many indirect space industry consequences of Russia annexing Crimea. They include stuff like export licenses being harder to get right now, companies in the US and Canada cancelling launches on Russian launchers, and Nasa saying that they're following sanctions while ignoring them completely.

There haven't been many direct Crimean things though, but a Russian-language newsletter suggest one: that Sea Launch may be relocated to Crimea.

Sea Launch is, as the name suggests, a company that launches satellites from an ocean platform. It was founded by Energia (Russian government space 'company'), Boeing (American company dependent on the US government), Aker Solutions (a Norweigan oil company with ocean platform experience), and a Ukranian (really ex-Soviet) company that actually builds the first and second stage of the rocket. Because that's not international enough, the two boats are registered in the libertarian paradise of Liberia.

It went through bankruptcy and is now 95% owned by Russia, but is still physically based out of Long Beach, USA, and launches in the Pacific. Because of that they have to get FAA launch licenses and deal with import/export/ITAR stuff.

Whether they'll continue to be able to operate as a Russian company with Ukranian first and second stages and American payload fairing and interstage structure is a big question by itself, but they're also talking about moving ports. Previously Vladivostok had been mentioned, but apparently Crimea is now in consideration since it already has the needed port facilities. It seems to me that it's a long ride from there to an open part of the equator, but Russia is not known for its low latitude deep water ports.

The article discussing this (in Russian) is on page 75 and 76 of this 8.5MB PDF: http://ebull.ru/dl/digest-075.pdf

I fixed all the line breaks and Google Translated it into English:

Roskosmos posted chapter RSC "Energia" last warning Chapter parent company manned spaceflight industry Vitaly Lopota may not be able to implement the idea Translation project "Sea Launch" from California to Crimea Roscosmos direct commission to verify RSC "Energia" - the parent company manned spaceflight industry. According to the head Oleg Ostapenko Space Agency, the Commission's task - to evaluate the production capacity of the enterprise to new tasks. Personal commission, according to Ostapenko, has not been determined. In most cases, verification of Roscosmos - a bad sign for the management of the auditee. During the previous inspection the same "energy" (spring 2013) were received, subsequently became the basis for the initiation of criminal proceedings against Vitaly Lopota, President of RSC "Energia". The main topic of the test was the project "Sea Launch"; since 2010, the system for launching missiles "Zenith" from a floating platform is maintained by subsidiaries of "energy", which are the beneficiaries of the Swiss Sea Launch AG, which owns the assets and "Sea Launch". The project is unprofitable because of the small number of orders for launches, and "Energy" keeps it afloat including credit facilities. During one of these operations, which resulted in the "Energy" has incurred, Lopota and was recruited as a defendant in the article on the abuse of power. At this time the commission Roscosmos "Sea Launch" will not check notes Ostapenko: - "Sea Launch" - a commercial project "Energy", and our aim here is to evaluate the capabilities of the enterprise in the production plan. Vitaly Lopota meanwhile keeps trying to save the project "Sea Launch". A source close to the leadership RSC "Energia", the latest idea Lopota - rebase "Sea Launch" in the Crimea and in the same place to create the infrastructure for the maintenance of the complex. From Russia in the Crimea industry promised investments that, according to Lopota, could be directed at creating a database service for "Sea Launch". Lopota himself refused to comment on the idea of ​​relocation "Sea Launch" in Crimea, stressing that at the official level, it is not made out of such a proposal. Currently, "Sea Launch" based in California. Roscosmos leadership last year concluded that the service in the U.S. restricts the use of facilities for the Russian Federation. Then Crimea was not Russian, and Ostapenko offered to relocate to consider "Sea Launch" in the Far East - in the Soviet port harbor or Vladivostok. Government's decision on the matter has been taken. According to experts Roscosmos, for approximately $ 1 billion can be created in the new port-based infrastructure required: assembly and testing facility, store the rocket manufacturing plant of liquid oxygen and nitrogen base for storage of kerosene and other objects. Oleg Ostapenko refuses to comment on the prospects of the complex "Sea Launch". The decision on the future of the project will be taken at the government level. To date, according to the first deputy general director of "Yuzhmash" Sergei Voigt, the "Sea Launch" has two potential customer: one wants to launch a satellite in 2016, the second - as fast as "Pivdenmash" will produce a booster. It remains an open question about the fate of the Vitaly Lopota: according to some, it can shift as president RSC "Energia" later this year. Soon, being state-owned shares of the company will be transferred to the United Rocket and Space Corporation (ACCD), whose administration including decide on the management of the company. Intransigence and uncompromising Lopota, as well as his desire to keep unprofitable "Sea Launch" at any price, can not play in favor of the current head of "Energy". In ACCD neither confirmed nor denied the information on possible replacements in the RSC "Energia". - Staffing solutions for management of the enterprises included in the ACCD may be accepted. But only after the corporation becomes really control them, - said the director of the Department of Media and Information Policy Joint Rocket and Space Corporation Igor Burenkov. News 06/06/2014

Original Russian, but I've removed the line breaks and the hyphens to get Google to translate it. I may have screwed it up somewhere because of that.

Really really big.

After my post identifying the Falcon 9 propellant cloud as the UFO that was seen over much of the Indian Ocean, I received a few further emails.

Devin Edmonds sent me links to a couple more photos people had posted to Facebook. One of the clearest was by Tsiaro Ryan'Avo probably taken from Madagascar:

It shows much less motion blur than many of the photos of the propellant cloud, so I hoped that it would be possible to identify the stars in it and from that figure out how big the cloud was.

I submitted the image to nova.astrometry.net which can analyze an image and automatically figure out where in the sky the camera was pointing. It gave this cool output:

It also gave a KML file as the output, which I opened in Google Earth. There I measured the height and width of the cloud and converted it from RA/Dec to degrees. The average value was 6.672 degrees. For comparison, the moon averages 0.528 degrees in width as seen from the ground, meaning that the Falcon 9's propellant cloud was more than 12 times as wide as the full moon! No wonder so many superstitious people thought that the world was ending.

From one of the images I showed in the prior post about this, we know that the rocket stage was about 1495km above sea level when it flew over Madagascar. With that distance and the angular diameter of the cloud, we can do some trig and calculate how big the cloud was in space.

The kind of amazing answer: 174 kilometers (108 miles) wide! Really big; Réunion, where it was seen from, is only 63 kilometers long. Madagascar's widest east-west width is 560km, so the cloud was almost a third as wide as Madagascar.

I hope planetary nebula scientists have looked at the results, as this was likely the largest nebula simulation to be photographed in recent years.

Reddit user p0nette posted this image to Imgur and Reddit with the question "Saw this tonight in Reunion Island, what is it ??!"

Here's a full-size crop of the cloud:

If your geography is as rusty as mine was, Réunion is a French island in the Indian ocean off the east coast of Madagascar. I wondered if the image might be related to today's Falcon 9 launch.

When an upper stage doesn't work correctly in space, it often leaves an interesting display. The unburned propellant creates a cloud in space that can reflect sunlight and be visible over a huge area. A previous SpaceX launch caused reports of a UFO over Australia.

Could p0nette's object be the upper stage of the Falcon 9 that was launched this morning? I think it is.

In the press conference after the launch, Elon Musk said that SpaceX attempted a relight of upper stage but encountered an anomaly. So it certainly could have left a cloud of propellant floating in space.

One of the missions on the Falcon 9 was DANDE, the Drag and Atmospheric Neutral Density Explorer, designed and built here in Colorado. They posted "Two Line Elements" to their website after the launch, which give the orbital information of the satellite. While the satellite and upper stage will drift apart over time, early in their orbits they should be fairly close together.

With the orbit data we can rewind time using Alexandru Csete's software Gpredict and find the course that DANDE, and the upper stage, flew over. The rocket launched at 16:00 UTC, here it is just after launch at 16:09 UTC:

Here it is an hour and a minute after launch, after it flew south over Antarctica and then north up the other side of the world:

The yellow circle represents the area over which it could be seen above the horizon. From Réunion, it would have been almost directly overhead, and would have taken some time to cross the sky. Réunion's time zone is UTC+4, so it would have been 9:01 PM local time when it was highest in the sky.

I haven't yet gotten confirmation from the original poster as to when it passed and what direction it was traveling, but this seems like the most likely candidate for the source of the cloud. I'll update further when I find out more.

Existing solid modeling software for computer aided design (CAD) is designed from the mindset of making a part on a mill or a lathe, and this is almost wholly unsuitable to the style of 3D printing.

(This is possibly a draft, I may come back and add images later. Just wanted to get something down.)

The typical workflow to design a part in many CAD programs like Solidworks or Autocad is as follows:

Draw a basic shape, like a rectangle or circle.

Extrude that shape in a dimension to create a material blank.

Do further operations on that blank, like cutting pockets, cutting chamfers, drilling and tapping holes, etc. Operations are easiest to set up on planar faces, like drilling a hole straight through the part.

Do finite element analysis on the part to see if it is strong enough for expected loads.

Until all of the stresses in the part are low enough, tweak the sizes and operations at step three and re-run step four.

Send the part to the machinist, get quietly cursed because you put a small chamfered pocket in on an otherwise untouched face.

To machine the part on the mill is basically the same process: get a piece of material as close as possible to the bounding box established in step 2, then do the drilling, pocketing, tapping, whatever.

The price of a milled part is implied in the design process: the more faces of the blank that you do stuff to, the more it's going to cost. Either a mill with extra axes is needed, so that it can tilt the tool in various ways, or the piece of material has to be held in the machine in a variety of configurations.

The price of milling a part is then the price of the base material, which is typically cheap, plus the time to do all the setups times the machinist rate, plus the time to do all the milling times the machine rate. Because of this, simpler parts are cheaper and parts whose final volume are close to the original stock dimensions are cheaper.

A 1"x1"x1" cube is cheap, a cube of the same size with pockets in all six faces is significantly more expensive.

This is not at all the case with 3D printing. The price to print something is a small fixed cost plus the volume of the final part times the price of material and machine time. It mostly doesn't matter how complex the part is, just how much of the build material is used, which also sets how much machine time will be used.

(There is some deviation from this; part height is often more expensive than part width or depth. Also, with some processes like supported FDM, complex shapes will require more support material. But on average volume is the price driver, unlike traditional machining.)

The means that an optimum design for 3D printing is minimum volume, rather than the traditional milling optimum design of minimum complexity.

The big challenge here is that all of the mature software has years of history of being used to design stuff for processes like milling, lathing, stamping, sheet metal bending, tube bending, &c. They now are starting to have small allowances for printing like having a "send to 3D printer" option in the menu, but it's not making dramatic changes in the user interface.

What we need is a different kind of CAD software, as different from the existing software as 3D printing is from milling. The way I see it working is as follows:

Define the constraints of the part: it needs holes here and here to bolt between these two parts, and it needs to hold this much force in this direction. Essentially you'd be telling it "this is where the part can't be, and this is what it has to hold."

Possibly give it one or more starting ideas for material between the constraints.

Let the computer run FEA on the part, stopping and adding or removing material wherever the stresses are too high or too low.

Control-P Print.

You would end up with close to minimum volume parts, with complex, organic-looking shapes with near no flat surfaces. This is no coincidence: this process is basically the same as how our bodies work on bones. High stresses on bones results in our bodies making the bones bigger and stronger. Low stresses lead to the bone being broken down, which is why astronauts lose bone mass.

Solidworks has a bit of this capability; you can design a part traditionally and then tell it to iterate FEA while twiddling some dimensions. But it definitely won't do anything new, it can only do things like change the depth of a pocket, not put a new strut between two points.

I might be asking for too much here, but hopefully someone out there is already working on this software, and maybe in a decade or so it will be just as feature filled, buggy and expensive as CAD software is now.

On ARocket we were discussing the possibility of a rocket pump competition, and how the rules for such a competition would work. The general idea of pumps in rockets is that they increase the mass performance of the vehicle, and that is a major win. However, most amateurs and small rocket companies avoid them, due to their complexity. The prize would be to encourage people to make attempts at pumps.

My preferred and shortest rule requires the pump to be in a complete working rocket system, and no one liked that. My second rule is rather more complex, and goes as follows:

The winner of the prize is the first person to demonstrate a propellant simulant pump with 'total mass' equivalent to one third that of an equivalent pressure fed tank, pumping at a rate sufficient to feed an engine with thrust equal to 125% of the pumped propellant mass, with a 550psig pressure differential across the pump, and where 'total mass' is defined as Mp+ln((D+Mp)/Mp). Mp is the complete pump system hardware dry mass plus any pump propellant remaining at the end of pumping, D is the mass of pump propellant that is dumped overboard during pumping.

In simple terms, the pump system has to weigh a third as much as the tank it's replacing and deliver propellant at a useful rate.

The idea there is that the pump has to be a lot lighter than an equivalent tank to be worth the investment in pump development. I picked 1/3 fairly arbitrarily, and will have amended that number by the end of this discussion.

I built a spreadsheet to see what mass a pump would need to be under in order to win a competition with that rule. For this theoretical pump I assumed a high density storable propellant combination as follows:

Oxidizer: 98% hydrogen peroxide, density 1.45 g/cm^3

Fuel: RP-1, density 0.81 g/cm^3

Mixture ratio, O:F, by mass: 6

Mixture ratio, O:F, by volume: 3.352

Average total density: 1.303 g/cm^3

Operating pressure: 515 psia

Engine operating at Pa 1 atm, Pe 0.5 atm

Delivered SL Isp at ~95% efficiency: 237.1s

The assumed tankage has a 12:1 inner fineness ratio, reasonable for a small stage, and is a cylinder with hemispherical ends and a hemispherical common bulkhead. The pressure fed tank operates at 600psig, the pump fed tank is at 50psig. 6061-T6 is the assumed material, operating at 2/3rds of a 45ksi UTS (i.e. burst is 1.5x MEOP).

I punched all that into a spreadsheet to calculate the mass of both tanks, then divide the difference by 3, then figure out the flowrate needed from the thrust.

Here's the output of that spreadsheet all scrunched up, go here to see it normal: http://i.imgur.com/0TqXd.png

In reply to me on ARocket, John Shilling suggested the following guideline instead, based on the work of Robert Goddard:

Flow Rate: 40 gallons/minute combined, fuel and oxidizer

Output Pressure: 500 psig

Total System Weight: 10 pounds

My 1/3rd weight and 1.25x thrust rule was clearly way too lax, and gave a burn time of over 3 minutes, which is way too long for a small rocket. I revised the idea to 1/10th weight and 2x thrust, which gives the numbers in the red area on the right.

It then turns out (in green) that a 10.35 lb pump system should deliver 41.05 gallons per minute, right near Goddard's numbers. That gives a two minute burn on a rocket with 400kg of propellant, which is just over the 200,000 lbf*s total impulse limit for a FAA class 3 amateur rocket.

In conclusion, I think the rule amended to 1/10th mass and 2x thrust is a good starting point. I still think the somewhat arcane bit with the natural log is necessary, it makes it possible for different pump cycles to compete on a level playing field.

In which I rewrite the first half of the Gilligan's Isle theme to entertain my mom.

Update: take the poll online at http://shuttles.questionpro.com if you don't want to send an email.

The Space Shuttle program is coming to an end. Nasa is soliciting bids for the locations of the retired space shuttles; the candidate museum must have an indoor spot for the shuttle and pay $28.8 million to have it readied for display and transported to the location.

There are currently four space shuttles: Discovery, Atlantis and Endeavour flew into space, and Enterprise was used as an atmospheric test vehicle. Enterprise is currently at the Smithsonian National Air & Space Museum, and early reports from Nasa said that Discovery would go there and Enterprise would be free to go to another museum.

So where will the four space shuttles go? Below is a list of some of the approximately 20 museums who are in the running. The final decision is made by Nasa, specifically Administrator Bolden, but there is a lot of pressure from politicians and advocacy groups.

Send me an email at [email protected] by April 5th with your four guesses, from the list below or any other candidates you know of. Of the people who guess correctly we'll randomly select one winner to receive a 1.65" square section of Space Shuttle tile material manufactured for Columbia, but not flown in space. (Sadly, tiles flown in space are sent to the dump by Nasa, and they prosecute anyone who sells them.)

Pick the four locations you expect to get a shuttle, not the locations you wish would get one. One entry per person. Administrator Charlie Bolden is not eligible for the competition. The announcement is set for April 12th, the 30th anniversary of the first shuttle spaceflight. The prize winner will be announced shortly after.

Smithsonian National Air and Space Museum, Steven F. Udvar-Hazy Center in Chantilly, Virginia

Kennedy Space Center Visitor Complex in Florida

Johnson Space Center in Houston, Texas

National Museum of the United States Air Force in Dayton, Ohio

Intrepid Sea, Air and Space Museum in New York, New York

The Museum of Flight in Seattle, Washington

U.S. Space and Rocket Center in Huntsville, Alabama

Adler Planetarium in Chicago, Illinois

California Science Center in Los Angeles, California

Brazos Valley Museum of Natural History in Bryan, Texas

Evergreen Aviation & Space Museum in McMinnville, Oregon

Tulsa Air and Space Museum & Planetarium in Tulsa, Oklahoma

March Field Air Museum in Riverside, California

Big thanks to Robert Pearlman of collectSPACE for adding a couple more entries to this list; you can see some museums' plans on his site, Photo Gallery: How to display a retired space shuttle.

Again, pick the four locations you expect to get a shuttle and send them to me an email at [email protected] by April 5th. I won't use your email for anything except to contact you if you won. If you'd like to be contacted with the result whatever it may be, mention that in the email.

You don't have to guess which shuttle will go to which location, just which locations get a shuttle. But if you do guess which goes where and get them all right, I'll find something extra to give you.

In late August of 2007 I built this catamaran rowboat out of construction supplies and rowed it across the Mississippi river and back. It was an extrapolation from boats of similar construction on Instructables, but built as a reasonably fast and very stable rowing catamaran.

Shortly after building it I moved away from Minnesota, to Mojave, California. It didn't find much use there, being the desert and all.

The boat was unfortunately forgotten in a storage room in Mojave when I moved from there to Dallas, where I live by a lake and could use it every week. I don't have a way to get it here, so it is free to anyone who wants it, they just have to pick it up in Mojave. Contact me asap: wikkit at gmail dot com.

The PVC tubes that made the load frames of the catamarans.

A complete frame, the two are identical. Styrofoam panels were added to make the loaded skin.

The complete boat on the ground. Tarp was added to make it watertight, like a stretched skin canoe. The sharp hulls make it track well, even though it's very short. It's under the length that requires boat registration in Minnesota and likely most states.

Overview. The deck is just wood. The boat no longer includes the oar uprights and oarlocks, I took them off for transportation to Mojave (upside down on the top of my van) and they've since been tossed.

Inner detail. Gussets were epoxied into the ends of each pontoon where the outer foam wasn't completely supported. The tarps were Harbor Freight, cut up and re-riveted so that they could be roped on.

Me standing on the boat floating free in the Mississippi. It's very stable due to the catamaran design, much more than a canoe.

Free to the first person who wants it. Need to pick it up soon, as the folks want their storage space. It's in Mojave, California, 93501.