Rocket Man Blog

Operational Costs

Back when I was in college I had an argument with a professor over a question he asked on a test. The question was “What is the number one factor affecting the cost of parts.”

I answered “Labor,” but he thought the answer should be “Quantity.” I tried to explain to him that while it is (almost always) true that the cost per part decreases as the quantity of parts fabricated increases, the reason the cost goes down is that less labor is consumed in making each part.

This not only applies to the labor the fabricator puts into making the parts, but it also applies to the labor expended at every step in the entire process. The time it took the engineer to design the part, the time the procurement person spent procuring the raw material, the time the stock room personnel spent stocking the material and delivering it to the floor and so on. As I explained in my post $500 Hammers, when you only make a single part, the material cost for the part is dwarfed by the labor cost, which is why a hammer could easily end up costing $500 if you only order one.

Fabricating a larger quantity of parts allows you to automate a large portion of the production process, thus decreasing the cost per part. Even the cost of the machine used to automate the process is driven by labor costs. A machine that takes less time to design, fabricate and maintain, or in other words has a lower labor cost per hour of operation, will decrease the cost per part versus using a machine that is poorly designed/built and thus more expensive.

A perfect example of this can be seen in the automotive industry. You can currently go out and buy a well built automobile for around $20,000. However if that same car was one of a kind, it would probably cost at least 10 times that amount to have it built, and the vast majority of that would be labor costs.

The next two drivers affecting the cost of parts are overhead and the cost of any fixed infrastructure required.

Overhead consists of any personnel not directly involved in the process of making parts, like secretaries, human resource personnel, accountants, lawyers, etc. and any equipment not directly used in making parts like copiers, printers, phones, office furniture, office space, etc. and fixed infrastructure is any item that is directly involved in making parts, like fabrication shops, machine tools, stock rooms, assembly lines, etc.

So the way to reduce the cost of an individual part is to minimize the amount of labor required, minimize overhead and minimize the amount of fixed infrastructure required. There are of course other things that affect costs like material, the amount of power required to make each part, etc. but unless the labor/overhead/fixed infrastructure cost per part is extremely low, they generally are not the significant drivers in the cost per part.

There are of course numerous exceptions to these general guidelines, but they hold true in most instances. In fact, anybody familiar with the business world already knows everything I have just described, and that term for the summation of labor, overhead and fixed infrastructure costs is called the Fully Loaded Labor Cost.

The reason I have provided this (extremely) short business primer is because all of the above discussion applies to processes as well as to parts. For example, some people have a hard time understanding why it costs so much money to launch the Space Shuttle. Well, if you divide the Fully Loaded Labor Cost of everybody working on the Shuttle program by the number of Shuttle launches per year, you will pretty much get the cost of each launch.

In short, the reason the Space Shuttle is so expensive to operate is because NASA has an enormous amount of people assigned to the program, their overhead is outrageous and they have a large amount of expensive fixed infrastructure to maintain. Consequently, even though NASA does not have to build a new vehicle for every launch like Lockheed Martin, Boeing, Arian, etc do with their expendable vehicles; having a reusable launch vehicle has not brought the cost of space access down because the total cost of operating the vehicle is so expensive.

So what would it take to reduce the cost of operating a reusable launch vehicle? The answer is to minimize operational costs (or OPS costs) by minimizing the number of people it takes to service the vehicle between flights, having a company organization with low overhead costs and minimizing the amount of fixed infrastructure required to launch and service the vehicle.

That is where the name of my new company comes in.

For those of you who don’t know, I recently went to work as the Lead Propulsion Engineer for TGV Rockets in Norman, Oklahoma. TGV stands for Two Guys and a Van, and it embodies the philosophy we are trying to design our vehicle to.

One day, Earl Renaud and Pat Bahn (The COO and CEO of TGV Rockets respectively) were talking about why the Shuttle was so expensive to operate, and (if I remember the story correctly) Pat made the statement that if they could operate a vehicle with two guys and a van, the cost per flight could be reduced significantly and the name stuck.

We sometimes joke that TGV could stand for Two Guys and a Van, or Twelve Guys and a Van or Thirty Guys and a Van, but the goal is the same: Design, fabricate and operate a vehicle so that the OPS cost per flight is as low as we can safely make it.

For the vehicle we are working on, the biggest cost per flight will be insurance (more on that in a future post), then OPS cost and then far below that will be the cost of fuel. Until we get a large number of flights under our belt the insurance costs will be something we won’t have a large amount of control over, but by minimizing the operational cost of the vehicle we feel that we can design, build and fly a fully reusable vehicle that will have a total price per flight of around $1 million excluding insurance.

In future posts I plan on discussing in more detail exactly what our design philosophy is, what we are currently working on and exactly how our design is progressing, but I will also continue to write about general space related issues.

I hope I can provide an entertaining and informative record of the process over the next couple of years, so if you have ever been interested in the process involved in taking a launch vehicle from conceptual design to first flight, stick around.

P.S. According to the ground rules I worked out with Pat and Earl, the things I cannot discuss are only limited to who our customers are, where and how we get our funding and who our contractors and subcontractors are. Everything else is open for discussion.

In Oklahoma

It has been a busy couple of months, but after an extremely long hiatus I finally have the time to resume blogging. I have moved into my new house in Oklahoma, mostly finished unpacking, settled into my new job and actually have some free time.

A lot has happened in the last couple of months, most of it mundane and boring, but I will give the highlights for anybody who is interested.

I started at TGV Rockets on March 29th, but my wife and three kids were still in Colorado waiting for our house to sell. We finally sold it on April 20th and I flew out to Colorado on April 23rd. The movers packed up our house on April 24th and we started the long drive out to Oklahoma. We arrived here on April 25th and spent four nights in the Residence Inn waiting for our stuff to be delivered. Finally on April 29th (which just happened to be my 40th birthday) we moved into our new house.

The rest of the time I have spent unpacking, hanging shelves, organizing, decorating, getting bank accounts set up and the millions of other things you need to do when you move to a new state.

Earl, Pat and I have set up some guidelines about what I can and can’t blog about work, and in future posts I will bring everybody up to speed just what it is we are trying to accomplish at TGV Rockets.

It is good to be back.

P.S. Sorry if I missed anybodies emails over the last couple of months. I downloaded over 700 messages to my computer before I closed my old email accounts and have not had the time to go thru them all yet. Also, my new email address is rocketmanblog _at_ sbcglobal _dot_ net if you want to send me something (it is also listed over on the right side of the page under email).

Definitely Not The Way Most People Get New Jobs

The last couple of weeks have been really busy and I have found myself to beat to write anything. But tonight I have the kids in bed, the house is clean (no small feat when you have three kids) and I am forcing myself to actually write something.

A number of people have expressed an interest as to how I ended up accepting a job with TGV Rockets. Well it all started out when I read an article from the New York Times about John Carmak and his vehicle the Black Armadillo. At the time I did not know a whole lot about Armadillo Aerospace and I ended up writing an article titled Fools And Their Money Risk Being Blown Up which was based exclusively on the Times article.

Unfortunately for me, the Times article was very poorly written and made John’s efforts sound amateurish and haphazard, so my post ended up being pretty critical of the Black Armadillo (a note to all bloggers: Never trust the New York Times). A few days later while I was doing some research about the X Prize, I read a lot more about Armadillo Aerospace and decided two things. First that I really should write another article about the Black Armadillo that was more representative of the serious effort John was conducting and second that it might be fun to write reviews of all the X Prize vehicles.

However, before I got a chance to write a more balanced article about the Black Armadillo, John emailed me. John obviously has a good sense of humor because the subject of his email was “From the fool,” and in the email he asked if I had read Armadillo Aerospace’s website before I wrote my original article. I replied that I had not, but that I had read it after I wrote the article. I also told John that I planned on writing another article that would be more representative of his efforts and I asked if he would let me interview him for it. To my surprise he agreed to my request and graciously answered a long list of questions, which allowed me to write my first X Prize review titled The Black Armadillo.

Since John had been nice enough to answer my questions, I thought maybe some of the other X Prize competitors would agree to let me interview them as well. So I picked a few competitors that looked interesting and emailed them some basic info about my website and my intentions and asked if they could spare the time to answer some questions. Again, to my surprise, I got almost unanimous positive responses from the competitors I contacted and I decided to write my next review about Bristol Spaceplanes Ascender vehicle primarily because David Ashford was the first person to respond to my email.

Since Pat Bahn from TGV Rockets was the second person to respond to my email, I decided to write my third review about The MICHELLE-B. After I finished my email interview with Pat, I thanked him for his time, wrote my review and told him that if he had any comments I would post them as an update to my article. Pat had a few minor corrections and clarifications that he emailed me after reading my article, and he concluded the email with the line “Are you any good, we are hiring.”

Since I was not really looking for a job at the time (which I told him) I sent him a quick summary of my experience and qualifications and figured that was the end of it. However, Pat had a few technical questions for me that I was more than happy to answer, which resulted in us exchanging a series of emails. During that time, Elon Musk from SpaceX read one of my articles that Clark Lindsey from HobbySpace linked to and Elon sent me an email asking if I would like to interview with SpaceX.

Since I am originally from California and would not mind moving out there again someday, I found Elon’s offer intriguing. So I did some research about SpaceX, which included emailing Pat and asking him what his thoughts were about them. Pat was again nice enough to help me out, but he also told me that I could call him if I wanted to talk about it in more detail.

I took Pat up on his offer and we talked for some time. During the conversation I also spoke to Earl Renaud for the first time and Earl asked if I would be interested in coming out to Oklahoma for an interview. I again told them that I was not really looking for a job, but that I would be interested in hearing what they had to offer if they really wanted me to come out there.

It took some time for the interview to be arranged, but I finally went out there a couple of weeks ago. And as of March 29th, I will be working for TGV Rockets as the Lead Propulsion Engineer.

TGV Rockets

As some of my readers already know, and I'm sure some have guessed, I have accepted the position of lead propulsion engineer with TGV Rockets. I gave my two weeks notice at my current job this morning and I will be starting at TGV on March 29th.

The last couple of weeks have been filled with getting my house in Colorado ready to be put on the market and going out to Oklahoma to find a new house, which is why I have not had any time to post.

I am really looking forward to the new challenge and I will fill you in on the details of the transition in a follow on post.

To Busy To Post

I feel like a broken record, but I again apologize for the lack of posting. I have been really busy lately with some big changes in the Oakley household, so posting will probably not resume for about another week (although if I do manage to find some free time I will try to write at least something).

I will fill everybody in about what the changes are on March 11th.

SpaceX

Sorry again about the lack of posting, but I just returned from a trip to California where I met with Elon Musk and the staff at SpaceX.

Elon seems to be a great guy and everybody I met with at SpaceX seemed intelligent and dedicated. It looks like he has built a really good team, which bodes well for the success of the Falcon Is maiden flight in May of this year.

I also saw their development vehicle and my first thought was “there is no way that vehicle will make it to orbit.” Not because of the construction, because the vehicle itself actually looked really good, but because it is so much smaller than the vehicles I am used to working on. The whole thing is about the length of just the LOX tank on the Atlas V, but it is much smaller in diameter.

Unfortunately, as with my visit to TGV Rockets a couple of weeks ago, I am not at liberty to discuss what we talked about. But I will say that after meeting with Elon and his team, I think there is a high probability the Falcon I will actually make it into orbit.

“Off The Shelf” To Orbit Doesn’t Exist

I frequently hear comments, both here and on other web sites, from people who think it would be relatively easy to build a launch vehicle by cobbling together components from other vehicles. The ideas usually go something like –

Just take a shuttle tank, put some RL-10 engines on the bottom of it, add a fairing on top of it and you have yourself a launch vehicle.

Or-

Just strap seven Atlas V first stages together, put 3 Centaur upper stages on top of them and you can carry over 100,000 lbs to orbit!

Or any number of other wild ideas.

What people who make these types of suggestions do not understand is that every component on an orbital launch vehicle is designed and built to perform a specific function. Since the structural mass of all orbital launch vehicles is such a small fraction of the liftoff mass of the vehicle, there is not a lot of margin built into any of the components, which means that if you change the vehicles configuration even slightly, you have to analyze, redesign and/or test the components to meet the new requirements.

To illustrate this I will give an example.

One of the current expendable launch vehicles is being redesigned for a heavy lift configuration, which means it will be able to carry a heavier payload. This increase in capacity is being accomplished by strapping two additional first stage vehicles to the sides of an existing vehicle, so you have three first stages in line with the upper stage and the payload on top of the center stage. Easy right?

Wrong, because the existing first stage was not designed to carry a heavier payload, nor was it designed to take the loading from the two additional first stages strapped to its sides. So not only does the existing vehicle need to be analyzed, redesigned and/or tested so it can be used as the middle first stage, but it also need to be analyzed, redesigned and/or tested so it can be used as the outer two first stages, because again, the existing vehicle was also not designed to be strapped to the side of a middle first stage.

Now, instead of a vehicle that uses three existing first stages, you end up with a vehicle that uses a new stage designed to be used as a middle stage and another new stage designed to be used as the two outer stages. And neither of the two new stages is going to be exactly the same as the existing first stage.

Another comment I hear frequently is that we should just start building the Saturn V again if we need heavy lift capability. Again, this should be easy right?

Again, wrong, because even if the drawings still existed for the Saturn V (which I am pretty sure they don’t), the components used on the vehicle are no longer being produced. This not only includes the engines, tanks and other structural components, but also the valves, bellows, hoses, avionics and all the other ancillary components needed to build the vehicle. Recreating the drawings and tooling required to make all these components would be a Herculean task, and then you would still need to re-qualify all of them, which would take a tremendous amount of time and money.

So the bottom line is that if you want a vehicle that is significantly different from any of the launch vehicles being produced today, you will need to spend the time and money required to design, build and test it. While a few components could be used “off the shelf” (primarily engines), the overall vehicle will require a significant amount of time, money and new (or at the very least modified) components by the time the vehicle is ready to fly.

Oklahoma Sure Is FLAT

I am sitting in a hotel lobby in Norman Oklahoma using their "Business Center" computer (which is just slow dialup). I came out late yesterday, had a meeting today and will be going home tomorrow, which is the reason for the lack of posting.

A prize will be awarded to the first person who correctly identifies who I met with. The people at the meeting, my family and one other person (you know who you are) are not eligible. (Note: prize has no actual value).

Update: Made it home OK. Andrew Case was the (almost) immediate winner of the "prize" by guessing Pat Bahn. Unfortunately I did not get to meet Pat as he was away on other business, but I did get to meet Earl Renaud and Kent Ewing of TGV Rockets, who are both great guys.

Don't Reflexively Dismiss Hydrogen

In the comments to my post Its Never As Easy As It Sounds, I was left in the unenviable position of having to defend a LOX/Hydrogen vehicle over a LOX/Kerosene powered vehicle. I am on the record as saying I prefer kerosene vehicles over hydrogen vehicles, because kerosene is easier to work with, requires smaller tanks and is easier to design around than hydrogen.

There is one benefit to using hydrogen though. It has a higher specific impulse (ISP) than kerosene. (For those of you who don’t know, ISP can be defined as the pounds of thrust obtained for 1 second of operation of the engine from 1 pound of fuel, or more generally how efficient the engine is).

I was unaware that there are people out there who are fiercely opposed to the idea of using hydrogen powered vehicles, because some of the comments and emails I received seemed almost reflexively dismissive in nature. Now there are a lot of ways to design a vehicle that can deliver useful payloads into orbit, but the basic criterion for any launch vehicle is the same. Deliver the most mass to orbit for the lowest cost with the most reliable vehicle possible.

Unfortunately these are essentially mutually exclusive objectives. Optimizing for one of them will invariably lead to sacrifices with one or both of the others. But by using hydrogen with its higher ISP instead of kerosene, you can increase your payload mass and the mass of your vehicles structure, resulting in a safer vehicle and suffer only an increase in the fuel costs and vehicle design costs.

I have designed both hydrogen and kerosene (and LOX) systems used in the testing of launch vehicles, so I know what I am talking about here, and I do not reflexively dismiss hydrogen just because it has some drawbacks. Being an engineer, my job is to determine the best possible design that meets all the relevant requirements, and not considering the most efficient fuel available because of some preconceived bias would be gross negligence on the part of any engineer.

Building An Actual Rocket

I have been meaning to link to this for some time, but Ted from Rocket Jones has been writing an excellent series of articles about the process involved in building a model rocket. Unlike me, Ted actually builds whole rockets, I just work on pieces of rockets. If you have ever wondered what it took to build a rocket, or built them yourself as a kid like I did, you should make sure to check out Ted's excellent series of posts titled Build It.

Searching For Something

I got quite a few hits this weekend from people looking for "rockets outer covering." Does anybody know why so many people were searching on that subject this weekend?

It’s Never As Easy As It Sounds

I enjoyed the first ever guest post on my blog, A Readers Rebuttal by Kelly Starks, and I want to thank him for spending the time to write it. Kelly’s article was prompted by his disagreement with my post Is Cheap Space Travel Around The Corner?, where I argued that until we have better engines, launch costs are not likely to come significantly down anytime soon.

Kelly disagreed, and in his article he presented a well thought out plan on how to make an affordable orbital vehicle that could drastically reduce launch costs below their current rate while still using existing technology. However, to build the vehicle he describes actually requires at least one better engine. Kelly describes the engine as follows –

Their zero- to maybe up to Mach-6 craft can be powered by modified versions of the engines that have been flying in the F-15's for 30 years. No need for new rocket-based combined cycle engines (though they might be nice). They just spray water into the engine intake to cool the incoming air to save the turbojets from overheating, and spray some liquid oxygen ahead of the burners to make up for the thin air.

I mentioned to Kelly that this concept could be considered a new engine, but his position was that it was a just a modification, not a new engine. Ideas like this are never as easy as they sound, and even according to the DARPA source Kelly cites in his article, this technology is still in the concept stage.

I worked on F-16 engines in my younger days as a member of the Air National Guard (they are the same engines that are used on F-15s) so I am pretty familiar with them. And while the development program to create the engine Kelly describes would not be as expensive as developing an entirely new rocket engine, it would not be cheap. So would it be considered a new engine or an existing engine? It’s definitely debatable.

The second engine Kelly would need for his vehicle is a rocket engine to power the flight out of the atmosphere. He mentions using a “RD-180'ish” engine in his article, but he does not specify exactly which engine he would use. Your choices of existing engines is limited if you want to use LOX/RP-1 for your propellants, and none of them have ever been used on a reusable vehicle. This not only means that they will suffer from some of the same problems Kelly points out about the Shuttle, namely components that are not easy to inspect/repair, but also that nobody is going to invest the kind of money it will take to develop the vehicle he is describing without doing some testing to determine just how reusable those engines are.

The RD-180 has been test fired a number of times successfully already, but it has never been subjected it to the kind of thermal cycling it will see if it is fired and then taken into the vacuum of space. I have stood on top of an RD-180 engine, and I know I would hesitate to make any determination as to how reusable it actually was without further testing.

So would an existing engine work for the vehicle Kelly is describing? Definitely, but the replacement rate is an unknown. An engine might last for the ten launches he describes, you might need to replace it every time or you might want to use a new engine that was designed from the ground up to be reusable and easy to maintain.

Later in the article Kelly presents some numbers as to how much each component of his proposed vehicle would weigh. He states that the vehicle would need 90 tons of RP-1 and 190 tons of LOX. Assuming his vehicle is using the RD-180 engine, he would actually need 245 tons of LOX (plus the LOX to run his modified jet engines), not 190 because the mixture ratio of the RD-180 engine is 2.72 to 1.

Now I’m not trying to nitpick his numbers here. I am always concerned that I will make a math error in one of my articles and somebody will derive great pleasure from pointing it out (fortunately this has only happened to me once before when I was calculating a pressure and accidentally used inches of water instead of feet of water and was off by an order of magnitude). In fact I pointed out this discrepancy to Kelly when he sent me a rough draft of his article, but I didn’t notice until after I posted his article that he hadn’t corrected it.

The reason I point this out is because Kelly states that his vehicle would use in-flight fueling to fill up the LOX tanks so it would not have to take off with all that extra weight. Unfortunately, no existing aerial tanker has the capacity to carry that much fuel, plus no existing tanker was designed to carry and transfer LOX. So, with the addition of the LOX that would inevitably boil off during the transfer, it would require at least two modified KC-10 Extenders to transfer that much fuel. That is assuming of course that all six of the fuel tanks on a KC-10 could be modified to carry LOX and that all the plumbing on the plane could be modified to transfer it.

Finally, even after designing and building a modified jet engine, choosing and testing an existing rocket engine and modifying a couple of fuel tankers, the mass fraction of the vehicle Kelly describes is almost exactly the same as if you had started on the ground using an engine that is just as efficient as the Space Shuttle Main Engines (almost exactly 85% propellent in both cases). However, you end up with the additional complication and weight of another propulsion system on the vehicle, which is one of the problems Kelly pointed out with the Shuttle.

With today’s better materials and manufacturing processes, I have no doubt that we can build an engine that is more efficient than the SSME. It would unfortunately have to be a LOX/LH2 engine, but the added complication of using liquid hydrogen would more than compensate for not having to use a second propulsion system or performing in flight refueling in my opinion. Plus, removing the jet engine weight would allow for a more robust vehicle structure, even after accounting for the heavier LH2 tanks.

While I like Kelly’s idea in principal, I think it is better suited for a two stage to orbit vehicle like the RASCAL vehicle DARPA describes (and the one Kelly links to in his article) or the one Pioneer RcoketPlane is proposing. Both of these vehicles would use a jet engine/rocket powered vehicle to get to the edge of space and then release a second stage to make its way into orbit.

In conclusion, I still believe that to achieve truly cheap space access you need better engines than the ones we have today. Because until space travel becomes as reliable and routine as air travel is today it will continue to be expensive. I do not believe this reliability will be achieved by using a vehicle that needs two propulsion systems and has to make two in-flight refueling stops. Instead it will require a vehicle that is as simple as we can possibly make it, and one that uses a single high ISP engine throughout the vehicles entire flight envelope.

Update: Kelly responds in the comments.

eTALKINGHEAD

Two bloggers who I read on a daily basis have been asked to join a group blog called eTALKINGHEAD . The blog is described as -

Political commentary, analysis and opinion on important events of the day :: eTALKINGHEAD.com is a multi-author political blog.

I had never heard of eTALKINGHEAD before, but with writers like Darren Copeland from Colorado Conservative and Stephen Maklin From Hold The Mayo contributing, it should be well worth reading. I would like to extend my congratualtions to both of them.

A Readers Rebuttal

Kelly Starks (a regular reader and frequent commenter) and I recently had an email discussion regarding my post Is cheap space travel around the corner? Kelly had some interesting arguments to make and I invited him to write a rebuttal to my post if he was interested. He told me he was definitely interested, and sent me the following well thought out article. (Note: The only changes I have made from Kelly's original article is to add proper formatting for his links.)

I recently found this Blog site and generally loved it (which isn’t really surprising since I’m an ex-NASA contractor and serious space buff). However, I did take issue with something Mark said in his Sep 13 article: Is cheap space travel around the corner?

Mark makes the statement that without new engines, low cost access to space is impossible. I responded in an e-let that, excluding some reliability and maintainability concerns, engines aren’t currently a significant roadblock to cheap, reusable space travel. Mark thought what I was saying would be worth a post, and invited me to make one.

I started my software/systems engineering career at the Johnson Space Center in the shuttle flight planning offices and mission control support operation. In the late 1980’s I moved to Reston, Virginia and worked on the Space Station Freedom program for a few years before winding up at NASA HQ in the office of space access technology. Almost all that time I worked for McDonnell Douglass, and had some friends involved in other programs such as SDI/DARPA's DC-X project. So I know a little about this. Don’t blow me off out of hand.

One of the things I like about DARPA's launcher projects is that they develop breakthrough systems by adapting off-the-shelf equipment. The DC-X demonstrated likely cost reductions of at least a factor of ten over the shuttle (i.e., costs of less than $1,000 per pound of cargo to orbit) while flying only as often as the shuttle. Of more interest is that folks inside McDonnell Douglass’ DC-X program privately assured me that frequent fleet operations might get costs to orbit down to perhaps $200 per pound of cargo. They did that using RL-10 engines first marketed in the ‘50’s, avionics left over from an airliner, and a bunch of parts so off-the-shelf that some were out of a junkyard! (The condensation trays were hubcaps from an old Toyota connected to a length of garden hose.)

How could such old equipment make such breakthroughs in costs, and how could a subscale, very sub-orbital craft prove it had done so? To be blunt, they studied what was expensive in the shuttle, and excluded it in designing the DC-X. G. Harry Stine covered this in detail, as well as a wealth of other items of interest to space buffs, in his book Halfway to Anywhere (there is a flight test photo of the DC-X on the cover). I highly recommend the book, but will briefly go over the pertinent point here.

The first thing you have to understand is what costs so much. NASA admits that the shuttles cost about $500 million to prep and launch. Others point out that NASA doesn’t count facilities and overhead costs that any commercial operator would need to bill for, but we’ll ignore that for the moment -- especially since most other launchers wouldn’t need those facilities. What costs in the shuttle are simple, basic, design issues. It was a MESS of a development program that forced a lot of sloppy design decisions. Those decisions made the shuttle a labor hog to prep for launch.

It takes about 10,000 technicians at Kennedy about four months to prep a shuttle for launch. That’s about $300 million dollars in labor cost out of NASA’s admitted $500 million expenses. That’s almost all due to unnecessary complexity (fragile tiles, multiple independent fuel and engine systems, reintegrating the craft per flight, etc). The DC-X took (and was required contractually to take) under 24 hours to service. On the subscale prototype this was accomplished with about ten people to service it. They calculated that it would take about a hundred people under a day to launch prep the full-sized production vehicle.

One of the major reasons it takes so much more time and labor to launch prep the shuttle is because it is actually three nearly-independent vehicles and five propulsion systems. After the external tank (ET) tore away during the Challenger disaster, the shuttle and two SRBs flew away independently. The two SRBs’ internal navigation systems realized they were on the wrong course after the breakup and steered them back on course. The exhausts in their wake created the fork-shaped cloud coming out of the big white steam cloud formed by all the fuel in the ET burning in midair. (The shuttle Challenger of course ran out of fuel and lost control, tumbled, and broke up soon afterwards.)

Because the SRBs are so completely autonomous, they need their own staffs to service them. Another staff mates them and the orbiter to the $30 million dollar ET. Several other staffs service the orbiter. The SRBs also fall into the ocean, and the seawater damage and the rest nearly totals them. Audits have found that it would be cheaper to simply let the SRBs sink into the ocean and buy replacements. They are not really reusable in a practical sense.

DC-X obviously eliminated all the servicing for the SRBs, and stacking the craft, but it’s not obvious what the rest of the cost savings are due to, which actually involves issues with the shuttle itself. Some of this is due to compromises caused by employing dated technology (computers, hydraulics, life support, etc.) that really should be replaced with better, more current technology. Maintaining the extremely fragile shuttle tiles is also expensive. A five year old could crush a shuttle tile in his or her fist. The tiles absorb water in the Florida air, so they must be Scotchguarded (yes, THAT Scotchguard!) between each flight. They can’t be glued to the shuttle’s old aluminum hull (NASA didn’t want to try those newfangled composites), so they are glued onto half-inch thick felt that's glued to the aluminum, then they are caulked to each other. Before each flight, each tile is carefully pull tested to make sure it hasn’t come loose, is checked for damage, recaulked, and then re-Scotchguarded.

DC-X would have been a bigger, lighter craft with more surface area. The effect is like a parachute. Each square foot has a far lower pressure load on reentry. The craft slows down faster at higher altitudes. Since it would not put as much load on the thermal protection system, it could use a simpler, harder system. The rocket engines could be less efficient and more robust. Actually, LOx/Kerosene engines would do as well. They and their kerosene tanks are so much smaller and lighter that they more than make up for the less powerful fuel. DC-X also didn’t have the separate hypergolic OMS propulsion and OCS system. It used small LOx / hydrogen engines like the main DC-X rockets. That eliminates another whole staff of OMS servicing personnel.

Beyond these points were a lot of very simple design faults, like service panels. Few were designed into the shuttle, so often it is necessary to dismantle whole areas of the shuttle to inspect or repair components. The result is like if your car didn’t have long metal pipes channeling dipsticks to the oil and transmission fluid. You would need to remove the engine every time you wanted to change the oil. They literally have to do things like that on the shuttle. DC-X’s design incorporated removable hatches; therefore, servicing tasks that took hours on other, similar craft took only minutes on the DC-X. It wasn’t really the technology that made the difference, but the design. This is the same way a Toyota is so much cheaper to operate than a Yugo or Trabant that uses similar technology.

Coincidentally, about the time I read Mark’s post I was reading online papers about DARPA’s newest low-cost launcher concept, and an in-flight LOx mining system by Andrews Space.

The Rascal papers are available at:
http://cism.jpl.nasa.gov/events/workshop/Preston_Carter.pdf

http://www.darpa.mil/TTO/rascal/RASCAL_PS_Final.pdf

http://hypersonic2002.aaaf.asso.fr/papers/17_5148.PDF

These sources show that the propulsion systems were far more advanced and capable than we space buffs had thought, which means that a lot of the technology issues we've been debating have long been solved.

DARPA has been doing tests on the engines the craft would need, and is designing the system around off-the-shelf parts. Their zer0- to maybe up to Mach-6 craft can be powered by modified versions of the engines that have been flying in the F-15's for 30 years. No need for new rocket-based combined cycle engines (though they might be nice). They just spray water into the engine intake to cool the incoming air to save the turbojets from overheating, and spray some liquid oxygen ahead of the burners to make up for the thin air. It’s like modifying a normal car engine for occasional drag racing. It has been increasing the thrust, without hurting the engines, for short bursts up to Mach-6. It will get the mother ship up to 200,000 feet or so, where it will drop off the upper stages from a bomb bay-like chamber. Those upper stages then boost up to the desired low or high earth orbits.

DARPA’s TSTO serves a wide range of small cargo needs from LEO to fairly high orbits. But if we do the math (fairly easy, given the off-the-shelf parts), it makes a black horse like HTOL SSTO possible -- or at the least an off-the-shelf HTOL TSTO. Also, Andrews Space is working on an LOx/liquid hydrogen system that can mine from the atmosphere all the LOx needed to get to orbit. They are proposing a TSTO that would cruise in the air to a launch point, filling its LOx tank en route. On the bad side their design is liquid hydrogen fueled, which makes it heavier and bulky. But the LOX mining gear was demonstrated in a lab, and got good extraction rates.

So some good news in the pipe.

http://www.andrews-space.com/en/corporate/projects(200311).html

http://www.andrews-space.com/en/corporate/NGLT(200311).html

I ran some numbers to see how this would work.

First consider a LOx/kerosene-fueled DC-X style SSTO: The fuel and LOx weighs 13 times as much as the rest of the ship combined. That’s doable according to the experts, and we’ve done expendable booster stages that could have been SSTO’s, but it’s damned hard to make a practical pure rocket SSTO. But of course we don’t need to.

Oh, by the way, I’m only considering kerosene as a fuel. It is not only easier to work with and cheaper than liquid hydrogen, but it makes for a far smaller and lighter craft. For example a LOx/hydrogen-fueled SSTO would only need nine times its dry weight in fuel, but for every eight tons of liquid hydrogen, the tank weighs a ton. One hundred tons of kerosene (or LOx for that matter) only needs a tank that weighs one ton. Even better, a winged vehicle can use the inside of the wings as the kerosene tank and eliminate virtually all the tank’s weight.

Anyway, we don't need just use rocket engines; we can use jet engines some or all of the way, which saves much of the takeoff weight. The bad news is that orbital speed is something like Mach 24, and past Mach 6 operating an aircraft in the atmosphere becomes a real pain, and speed is lost to drag. So getting out of the air at about Mach 6 is considered a good compromise. When I ran the numbers, when using jets like DARPA is working on to get to Mach 6, and then boosting to orbit with rockets, instead of needing 13 times the unfueled weight of the vehicle to get into orbit, it requires only about six times the dry weight. This is far more doable for the vehicle, and uses smaller engines.

To save more engine size and make a more convenient craft, use a winged craft like an airplane. Rockets need more thrust than their total weight to get off the pad, and then waste a lot of power accelerating straight up. Almost all aircraft have far less thrust then their weight (even an SR-71 has only half as much thrust as weight). So we can get by with fewer engines. Of course aircraft can’t normally take off with six times their weight in fuel and LOx. An SR-71 can only take off with a little over twice its dry weight in fuel. Two ways to get around that are to take off much faster, where the wings get far more lift, or fuel in midair. The latter is much easier on the airplane, and a LOT safer!

We can do normal mid-air refueling tricks like the military has been doing for decades, and add the fuel and liquid oxygen, or we can take off with all the fuel but not the liquid oxygen. This latter approach permits us to take off with fuel (kerosene) weighing about three times the dry weight of aircraft and add about the same weight of liquid oxygen during midair "refueling". Or we can carry an air liquefier to extract and liquefy oxygen from the air, like this rig:

http://www.andrews-space.com/en/corporate/Alchemist(200311).html.

Anyway, doing all this prepares for the boost to space in midair, without needing to use rockets until then. Just hit the jet engines and jet as high as possible until they flame out, coast for a while, and then kick in the rockets.

So, assuming you want to use off-the-shelf gear, you can do like DARPA, and take old F-15 engines and afterburners, and spray water and liquid oxygen into into the air intakes. The net result is a lot more thrust and speed for short bursts. So that gets the craft up to speeds like Mach 6 and arcing up to 200,000 feet of altitude where the rockets have no air to get in the way. You could then get the craft to orbit using something like a RD-180 rocket engine burning the fuel and LOx you carried to Mach 6.

So adding it up and skipping over the spreadsheet and equations, assuming a craft with cargo and everything it needs in space,weighs about 50 tons:

- It takes off with 90 tons of Kerosene. (A smaller fraction than the 1950's SR-71A could. We could also take off with less kerosene and add more in flight.)
- It carries modified F-15 jet engines that weigh in at 10-16 tons (depending on the acceleration rate you want), which is about the weight fraction of the engines in a F-15.
- The empty Lox tanks weigh under two tons, and the kerosene tank less than a ton. Note the Kerosene "tank" is likely mostly the wing, so it will likely weigh less.
- The rocket engines, like P&W;'s RD-180s, weigh under five tons, depending on the ascent trajectory. (The shuttle’s engine thrust is less than its weight after SSRB separation, but it gets to orbit either way.)

190 tons of LOx are added in flight, either with mid-air refueling or with mid-air oxygen mining and liquefaction. The plane can fly with this much weight as long as it’s going several hundred miles per hour, although it’s probably handling like a pig.

We can bring all the jet engines to full power and boost out for speed and altitude. When the engines finally flame out at Mach 6 , leaving all but wisps of the atmosphere, we start the RD-180'ish rocket engines. They consume the rest of that huge fuel reserve, bringing the 50 ton craft and cargo into orbit. Assuming heaviest assumptions for jet and rocket engine weights, everything else now weighs 27 tons. We can drop the cargo (likely 5 tons or less) or dock with a station. When we want to come down, we use a small burn to decelerate, and reenter.

Reentry can be a real strain on a craft. The smaller and heavier it is, and the faster it’s hitting, the worse it is. Orbiters are about 122.17 feet long with a wingspan of 78.06 feet and weigh up to 110 tons on reentry. SR-71's are 102 feet long with a 57 foot span and weigh about 27 tons unfueled. Our craft would need to take off with more than twice the load of an SR, and have big tanks for the LOx, and fly with a heavy load of fuel and rocket engines. So as a guess it would be under half again as long and wide, but have proportionally lighter structure and engines (some advances have been made in the last half century).

So we’re talking a craft with something like twice the surface area on reentry, and half the reentry weight of a shuttle orbiter. That should lower the heat energy per surface area down to a quarter that of the shuttle, possibly lowering the temps from the 1,200 and 2,300 degrees Fahrenheit the shuttle’s tiles face perhaps down to the 600-1100 degrees the SR-71's hull plates faced. If so, we don't need additional reentry material. If not, we can likely get by using both thick layers of rubbery paint that burn off, and high temp nickel-based alloys or carbon for the leading edges. Or we can use some other thermal control trick.

Using a water transpiration-based reentry system and some rough numbers I found at and , the weight of water needed for reentry cooling came out to about 10% of reentry weight.

After reentry we start a couple of the jet engines and fly to a nearby airport, or refuel from a waiting tanker and fly to a distant airport, and land normally. The shuttle faces far more reentry heat loads than it needs to so it can do hypersonic maneuvers to land over 1,000 miles to the side of its reentry track. That would be completely unnecessary for the craft we are talking about, which would also lower the reentry temp loads and allow less thermal protection.

The point of this is to save costs. We've eliminated most of the really expensive parts of the shuttle. Jet engines left over from '70's fighters are serviceable and last a while. The RD-180s are derived from an RD-170 the Russians designed and tested as capable of taking 20 flights. Pratt and Whitney lists the RD-180's as costing about 10 million and being reusable, but doesn't specify how reusable or how expensive it was to service them between fights. Assuming they were designed for servicing, and last at least 10 flights, they might cost a million dollars or so per flight (although if we fly a lot we can get cheaper prices from the manufacturer). Note: P&W;’s RL10 LOx/LH rockets are considered likely to last 300 flights with servicing of the turbopump bearings every 25 flights or so. Assuming no real exotic hull and systems checkout, current estimates for a RLV would suggest about 50 man days of labor, say $20,000 - $30,000 (compared to shuttle’s $300,000,000)? Fuel costs would be under $70,000.

Assuming all costs come to $4 million a flight, and it can lift 5 tons of cargo, that’s about $400 per pound of cargo to orbit, compared to the Shuttles and most expendables’ costs of well over $10,000 per pound to orbit.

One of the oldest serious proposals to build a craft like what we are talking about (though on a far larger scales) was Star-Raker by Rockwell in the '70's.
http://www.abo.fi/~mlindroo/SpaceLVs/Slides/sld047.html. It took off and landed from a runway with its fuel and oxygen load, and used rocket engines and ramjets for flight and boost to orbit. It was expected there would be little issue with reentry heating due to its large size and proportionally low reentry weight.

For examples of other past government RLV programs from '60's through early '90's, see:

http://www.abo.fi/~mlindroo/SpaceLVs/Slides/index.htm

VTVL
http://www.abo.fi/~mlindroo/SpaceLVs/Slides/sld043.htm

VTHL
http://www.abo.fi/~mlindroo/SpaceLVs/Slides/sld041.htm

Kelly Starks

Update: For my response to this article, see It’s Never As Easy As It Sounds

Giving Credit If Credit Is Due

While I was reading Rand Simberg’s post titled Oops, He Did It Again dated 1/29/04 where he deconstructs an article by Gregg Easterbrook titled Red Scare dated 1/22/04, I came upon this passage from Gregg’s article –

…while a Mars visit would be an exhilarating moment for human history, planning for Mars before improving space technology is putting the cart ahead of the horse. Nasa's urgent priority should be finding a new system of placing pounds into orbit: If there were some less costly, safer way to reach space than either the space shuttle or current rockets, then grand visions might become affordable.

Which sounded eerily similar to what I said in my post titled Putting The Cart Before The Horse dated 1/11/04 -

We shouldn’t put the cart before the horse and undertake missions using existing high cost expendable launch vehicles. Instead we should spend some money up front to significantly reduce launch costs.

It made me wonder if Gregg read my article before he wrote his. If so, it would have been nice if he had credited me in his article.

Let A Thousand Flowers Bloom

I have missed the opportunity to write about a lot of things in the time I was gone. Spirits troubles, Opportunities landing, the Space policy initiative proposed by President Bush and I’m sure a lot of other interesting things as well. But rather than repeat what I am sure others have written about all those, what I want to do is present my views on the future direction of spaceflight.

People have talked for a long time about the commercialization of space and how that would drive an increase in demand for space access. Space tourism, crystals grown in micro gravity, pharmaceutical manufacturing, mining the moon/asteroids and solar powered satellites are just some of the ideas that have been proposed for the uses of space.

However, space is already commercialized. Last year there were a total of 63 vehicles launched into space, and 24 (or 38%) of them were used to launch communication satellites into Earth orbit. Data is a little hard to come by, but there are currently hundreds of functioning satellites orbiting the Earth and a large percentage of them are privately owned and operated.

I think people who bemoan that space is not commercialized are actually complaining that space is not commercialized enough. And with the current cost of launching mass into orbit being so high, it is actually amazing anybody can make money there at all.

A lot of numbers are thrown around about how much it costs to get into orbit, but I will use a vehicle I am pretty familiar with to give you some idea as to what the approximate cost per pound of launching payloads into Earth orbit actually is. The basic Atlas V 401 can launch up to 27,500 lbs into low Earth orbit and it sells for around $90 million. So if the vehicles full capabilities were used, the cost per pound would be around $3,100.

Unfortunately, not all launches use a vehicles full capability, and with the cost per launch essentially fixed, if a payload is lighter than the maximum a vehicle can carry than the cost per pound goes up. Of course, a different launch vehicle can be used to launch lighter payloads, but the cost per pound would probably still be around a minimum of $3,000 (I know SpaceX plans to offer their vehicles for around $1,200/lb, but until they actually launch a few vehicles their actual cost/lb is an unknown).

Now it is not the purpose of this article to debate the cost/lb of every current and/or planned launch vehicle. I am merely pointing out the fact that it is currently very expensive to get anything into orbit. In fact, the 63 launches last year cost a total of over $4.3 billion, and in my opinion a lot of that money was wasted.

The reason I say the money was wasted is because it should not be so expensive to launch mass into orbit. There are a lot of technical challenges involved in getting a vehicle from the surface of the Earth into space (I know from experience because I deal with them on a regular basis), but there is no technological reason why the cost is so high.

The cost of flying cargo on airplanes started out quite high, but it came down over the years as better and better airplanes were built. Sadly, the same progression has not occurred in the area of launch vehicles. It’s as if we reached the technological level of the DC-3 airplane with launch vehicles and decided that was all we needed to make a viable industry. While the DC-3 arguably ushered in the era of worldwide air travel, it would not be possible to have the vibrant aviation industry we have today if we had stopped there. And while current launch vehicles have allowed us to achieve limited commercialization of space, it is not possible to have a vibrant space industry with the existing crop of launch vehicles we have today either.

In order for a vibrant space industry to develop we need a new launch vehicle that can drastically reduce the cost of space access from thousands of dollars per pound to at most a few hundred dollars a pound. Unfortunately there are two key obstacles standing in the way of this wholly achievable goal.

The first obstacle is the development cost. The cost to develop the necessary components for a low cost space transport, better engines, lightweight robust thermal protection systems and strong yet lightweight structural materials are all quite high and the only two entities with the resources to develop them, governments and large corporations, have not been very interested in spending the money it would take to make it happen. Even after all these components have been developed, the vehicle must still be designed, tested and built. Add these costs up, and it would not be cheap to develop such a vehicle, but I have no doubt that the long range cost savings would easily pay for development costs.

The second obstacle is all the special interests competing for the money in the space budget. Bob Zubrin wants to go to Mars, President Bush wants to send people back to the moon, scientists want Earth observation satellites and/or planetary probes to further their studies, etc. etc. etc. Now I have no problem with private corporations paying whatever they want to for space access. If they believe they can make money with current launch costs, more power to them. We also have a compelling need for spy satellites, so paying what it takes to get them into orbit makes sense.

But it is insane to pay the current launch rates to undertake all the other ancillary missions like manned spaceflight, Earth sciences or interplanetary exploration when the money could be better spent developing a vehicle that would make the cost of space access affordable.

The worst way to manage any endeavor is through the use of central planning. I always say that if central planning were the path to prosperity then the Soviet Union would have been the most prosperous country in the world. But central planning is exactly what we have in the space industry. The government decides what launch vehicles should be built, they decide what missions should be undertaken, they decide what should be researched, they decide who will be lucky enough to go into space.

I say let a thousand flowers bloom. If a group wants to go to the Moon, nobody should stop them. Same thing with Mars. If a company wants to mine the asteroids or build solar power satellites, great. If somebody wants to build a hotel in orbit, more power to them. Get the government out of the way and let people do what they want to in space and who knows what people will decide to do there.

But none of these things will be possible until the cost of space access comes down. As long as we continue to waste money launching missions that can wait a few years while we develop an affordable launch vehicle, the infighting over the limited space budget will continue. As an example, if you don’t think that we could have developed a low cost space transport for the eventual $100 billion total cost of the space station, than you are fooling yourself.

If you are one of the single issue space advocates, I urge you to change your issue to developing a low cost space transport. Because once the cost of space access comes down, the probability of your dream becoming reality will increase exponentially.

Back To Normal

Sorry about the long hiatus but things should be back to normal now and I should be able to resume blogging this week.

Swamped

I have been swamped this last week, both at work and home and have not been able to write anything. Blogging will resume soon though.

The New Space Policy Initiative

I just finished watching Bush's space policy speech, and it was pretty much as reported. I was disappointed by the lack of details; it seemed more like a political speech than anything else, and I guess we will have to wait and see how NASA turns the broad proposals Bush put forth into a concrete plan. I will write a more detailed summary later.

Update: Rand Simberg, who was not impressed by the speech, has more details and additional analysis.

Jay Manifold covers the Scope, Schedule and Resources.

Update 2: Cerdip is disappointed too.

Fred from The Eternal Golden Braid has the transcript of the speech.

A Message From Sean O'Keefe

I just received an email from Sean O'Keefe, the NASA Administrator, regarding the upcoming Bush space policy imitative -

I'm delighted to announce that President George W. Bush will visit NASA Headquarters Wednesday, January 14, to offer his vision for America's continued leadership in space exploration.

The President's remarks will be broadcast on NASA Television and on streaming video at our website (http://www.nasa.gov) beginning at 3 p.m. EST. All NASA Centers will be making special arrangements to televise the event in their auditoriums or other common areas.

NASA is always at its finest when faced with grand challenges. Tomorrow, the President will set in motion a wonderful opportunity to extend our exploration horizons and open new doors of possibilities for the next generation of explorers. Through our work to implement the President's vision, I'm confident we will energize U.S. creativity, innovation, technology development and leadership. I encourage all NASA and contractor employees to watch the President's remarks on what should be a very historic day for our storied Agency.

Sean O'Keefe
NASA Administrator

Breaking the Cycle

Wretchard from Belmont Club responds to my post Putting The Cart Before The Horse in an excellent post titled The Beckoning Sky –

Current launch costs are on the order of $8,000/lb, a number that will have to be reduced by a factor of ten for the habitation of the moon, the establishment of La Grange transfer stations or flights to Mars to be feasible. This will require technology, and perhaps even basic physics that does not even exist. Simply building bigger versions of the Saturn V will not work. That would be "like trying to upgrade Columbus’s Nina, Pinta, and Santa Maria with wings to speed up the Atlantic crossing time. A jet airliner is not a better sailing ship. It is a different thing entirely." The dream of settling Mars must await an unforeseen development. Although Rocketman hopes that President Bush will devote additional resources to developing better propulsions systems, he cannot produce this "different thing" to order.

I believe it is entirely possible to reduce launch costs to around $1,000/lb with incremental advancements of current technology. However, even if we decided to undertake a manned mission to Mars using existing launch vehicles at current launch costs, it would still be technologically feasible to accomplish, although it would probably not be economically feasible.

But I agree with Wretcherds observation that for permanent “habitation” of the moon or Mars to be feasible, we will almost certainly need new technology “that does not even exist” today. The problem as I see it is that with the current cost of space access being so expensive, the demand for that access pretty low. And since demand is so low, the amount of money spent on basic research into how to reduce the cost of space access is also small. In short, without more demand the cost will not come down, but the demand will not go up without the cost first coming down.

This cycle can be broken by either an “unforeseen development” as Wretchard points out, or by artificially increasing launch demand. A program of manned exploration is one way to increase this demand, but another way was proposed by Rand Simberg in his post A Quality All Its Own –

If the government, whether ours or Japan's, wants and needs assured access to space (and both must clearly think they do, because they continue to spend and perhaps misspend billions of dollars on it), it will have to decide to buy more than it thinks it needs to ensure that it has what it needs at an acceptable price. The decision makers must consider the possibility of simply putting out an order for currently-unthinkable numbers of launches and pounds of payload to orbit, to allow the private sector to do what it does best--driving down costs and increasing quality through competition and volume.

As the cost of space access comes down, uses for that access that we cannot even imagine today will be thought of. Hopefully, this will create a self sustaining cycle where lower launch costs will lead to an increase in demand that will lead to even lower costs, etc. And once space travel becomes routine, the odds of an “unforeseen development" happening that will allow humans to start colonizing other planets will greatly increase.

As Wretchard said so eloquently -

Clearly the day will come when nations will expand beyond the confines of the planet and our task is to be ready to mount the first real breeze for the distant shore.

Putting The Cart Before The Horse

It has been widely reported that President Bush will announce a new space policy initiative later this week that will include sending humans back to the moon and eventually on to Mars.

The visionary new space plan would be the most ambitious project entrusted to the National Aeronautics and Space Administration since the Apollo moon landings of three decades ago. It commits the United States to an aggressive and far-reaching mission that holds interplanetary space as the human race's new frontier.

Regular readers of my blog know that I am a strong advocate for manned Mars missions, but that I am also deeply suspicious of NASA’s ability to effectively administer a large scale program of this type. This is no longer the 1960’s, the cold war is long over, and a large portion of American public will not tolerate the level of spending it took to accomplish the Apollo program, especially if, as with most modern-day NASA programs, it ends up being significantly over budget and behind schedule.

Of course nobody has heard the actual proposal yet, and I am waiting to hear exactly what will be proposed before I pass final judgment on it. But in my opinion there is one item that absolutely must be included as part of the proposal, and that is a new launch vehicle that will drastically reduce the cost of space access.

The most conservative scheme for the amount of mass that must be launched into low Earth orbit for a manned Mars mission is probably the Mars Direct plan. Mars direct assumes that all the fuel required for the return trip will be manufactured on Mars, but even with this weight savings the plan still requires launching 200,000 lbs into LEO. With current launch costs being around $8,000/lb, just the launch costs for this plan would add up to $1.6 billion.

Mars direct also assumes that it would only take two launches to accomplish the mission, but the problem is that no current launcher has the capability to launch that much mass at one time. The Saturn V had that capability, but they are of course no longer being produced. Recreating the Saturn V would be no easy feat, but it could be done. However, recreating a vehicle designed and built over 35 years ago is not a step forward, it is a step backward.

If the Bush proposal ends up being similar to what has been reported, it will require even more mass to be launched. For one, NASA is unlikely to undertake a mission as ambitious as Mars direct, but also because sending humans back to the moon as part of the plan will require even more launches. I don’t know how much total mass will need to be launched to fully realize the goals Bush will propose, but it will be significant.

But what if we could reduce launch costs down to $1,000/lb?

Let’s assume that by the time humans land on Mars, a total of 400,000 lbs has been launched into LEO. At current rates, launching this much mass would cost around $3.2 billion. But at $1,000/lb the cost would only be $0.4 billion, for a savings of around $2.8 billion.

And that is where the title of this post comes in. We shouldn’t put the cart before the horse and undertake missions using existing high cost expendable launch vehicles. Instead we should spend some money up front to significantly reduce launch costs. $2.8 billion is an enormous amount of money, and if we can design and build a fully reusable space transport that can reduce launch costs to $1,000/lb for that price, the overall cost of the mission would be exactly the same as if we used existing launchers.

None of the three main components used in the Apollo program (the Saturn V, the command module and the lunar lander) saw any significant use after the program ended, and it would be a tremendous waste to end up in the same situation after spending the kind of money it will take to land humans on the surface of Mars. Hopefully, the ultimate legacy of Bush’s proposal will be to truly open the solar system by dramatically reducing the cost of space access.

For other opinions on this proposal, see Rand Simberg, Jay Manifold and Chris Hall.

Hold The Mayo

Stephen Macklin's Hold The Mayo has moved off of blogspot. Go check out his new site.

A Necessary Complication

A reader writes -

The NASA website takes great pains to describe the descent status tones emitted by the mars explorer craft. They seem like a fairly obvious way to keep track of what's happening during a landing that you can't otherwise observe in person, yet the way NASA describes them, it sounds like they're a relatively new addition. Is that the case? Are they new? If so, it would seem to explain the uncertainty as to why previous attempts haven't worked.

Matt

The tones are a new addition for Spirit and Opportunity, and they were used because of the uncertainty as to why the last lander, Mars Polar Lander, failed. If either Spirit had failed or Opportunity does fail during its descents, the hope is that these signals will allow us to better understand what the failure was so we can fix it in future missions.

My lab was involved in the testing to determine what the most likely cause of the Mars Polar Lander failure was. We concluded that the sensor on the bottom of the landing legs, which was supposed to shut off the thrusters once the lander had make contact with the Martian surface, was almost certainly activated when the legs were deployed, prematurely shutting off the thrusters and causing the lander to free-fall to the surface. However, even though this was the most likely failure mode, we are not absolutely certain it was what actually caused the mission to fail, or even if there was more than one failure.

Since I work in a test lab, I love to have all the data I can get about what is happening to a piece of hardware. The more data you have, the better you can make your next design. But being an engineer, I also like simplicity, because the simpler something is, the easier it is to design, build and test, and the less likely it is to fail during use.

Building hardware that can successfully make it all the way to the surface of Mars is a tremendous engineering challenge, and to decrease the chances of mission failures, past landers have been kept as simple as possible. Plus, the more hardware that is added to a lander, the heavier it becomes, which further complicates the design.

But adding the ability for a lander to communicate what is happening during its descent is a complication that is well worth including. Because hopefully we will continue to send landers (eventually with humans on board) to the surface of Mars, and the more data we collect about what happens during a Mars landing, the greater the chances of success for future missions will be.

Belmont Club

I want to thank Wretchard from Belmont Club for the link. It is always nice when somebody whose writing I respect thinks enough of something I have written to link to it. If you have never visited Wretchard’s site, I highly recommend you do.

How Spirit got to Mars

I was watching the NASA channel the other night and they had an animation of the Spirit mission starting from launch and ending as the rover drives off to explore Mars. I did not work on either the Spirit or Opportunity rovers, but both missions are similar enough to others I have worked on (Pathfinder and Mars Polar lander) that I knew what was happening during the animation even though there was no commentary to go along with it.

For those of you who have not seen the animation, hopefully my description of the mission will give you a better appreciation of just what it took for the Spirit rover to successfully make it to Mars. And for those of you who have seen the animation I am talking about, it might fill you in on some of the things that were shown.

Spirit was launched aboard a Delta II 7925 from Cape Canaveral on June 10, 2003. The Delta II consisted of 3 separate stages and 9 solid rocket motors. The launch vehicle weighed almost 527,000 lbs and on top of it, enclosed inside the payload fairing, was the Spacecraft. The entire spacecraft weighed a total of 2,343 lbs, but the Spirit rover weighed only 408 lbs. So for every 1 lb of rover that would make it to the surface of Mars, 1,296 lbs of launch vehicle and spacecraft were required to be launched from the surface of the Earth.

The enormous amount of energy required to get landers all the way to the Martian surface imposes severe restrictions on the design of the vehicles, which is why it is no easy feat to have a successful mission. There are a lot of critical systems that have absolutely no backup, and if one of them fails, the whole mission will fail. Which is why every component, system and subsystem on a planetary probe is rigorously tested prior to launch, and it is also one of the reasons why these types of missions cost a lot of money.

But when all those critical systems work perfectly you get spectacular data, as these pictures from the Spirit rover show.

Spirits mission started when the first stage and 6 of the 9 solids on the Delta II were ignited on the ground, followed by the remaining 3 solids one minute into the flight. The solids burned until they were exhausted, and then the first 6 solids were jettisoned, followed a minute later by the remaining 3. The first stage continued to burn until all its fuel was exhausted and then it was dropped back to Earth.

Next, the second stage was fired to place the spacecraft into low Earth orbit, and then the payload fairing was jettisoned. Once the proper orbit was achieved, the second stage engine was shut off and the spacecraft coasted in this orbit until it arrived at the position that lined it up on the correct path to Mars. At this point, the second stage was reignited and the spacecraft was placed on (almost) the proper trajectory towards Mars. Once the second stages fuel was exhausted it was also jettisoned, leaving only the third stage and the spacecraft to go on.

The third stage of the Delta II is pretty simple and does not have any axis-control systems. So to make sure the rocket stays on course while the third stage engine fires, four thrusters located around the base of the stage are fired to spin it up to 70 RPM. This spin helps to stabilize the stage exactly like the rifling in a firearm stabilizes a bullet during its flight. After the proper spin is achieved, the third stage engine fires until all its fuel is exhausted, and then two counter weights on tethers are released from the base of the stage to slow the spin down to 12 RPM much like a spinning ice skater slows down their spin by extending their arms.

Finally, with the launch vehicles job completed, the spacecraft was released from the third stage and it continued the journey to Mars on its own. However, it was not on the right trajectory to land on Mars yet. As I discussed in my post Life on Mars: Human or Martian, NASA does not want to accidentally contaminate Mars by having an upper stage crash into the planet, so after spacecraft are released from their final stage they have to make a course correction maneuver to place themselves on the proper trajectory.

At this point, the spacecraft consisted of two parts. The cruise stage that would control/power the spacecraft and allow it to communicate with Earth during its seven month cruise to Mars, and the entry, descent, and landing system that would actually protect the rover as it entered the Martian atmosphere. The rover’s batteries were fully charged prior to the release of the cruise stage, but unless the solar panels on the rover could be deployed soon after landing, the mission would have failed.

Once the spacecraft reached Mars, the cruise stage was jettisoned and the second part of the spacecraft entered the Martian atmosphere. The entry, descent, and landing system consisted of a heat shield on the bottom of the craft and a conical backshell to fully enclose the lander. Inside the lander was the rover, and on top of the backshell was the parachute and the rocket assisted descent (RAD) motors.

The heat shield protected the craft from the heat generated as it slowed down in the Martian atmosphere. Once it had slowed down sufficiently, the parachute was deployed and then the heat shield was jettisoned. After the heat shield had safely cleared the craft, the lander with the rover inside was lowered away from the backshell on a long tether. The craft then floated down on the parachute until the onboard radar system sensed it was close to the ground, the airbags were inflated and then the RAD motors were fired to further slow the craft.

Before the RAD motors had finished firing the tether was cut to release the lander. This was done so that the backshell/parachute would not interfere with the lander as it reached the surface of Mars. The backshell with the RAD motors still firing shot away from the lander as it headed towards surface of Mars. The airbags cushioned the impact and the lander bounced and rolled for quite a distance before it finally came to a stop.

After the lander stopped, the airbags were deflated and retracted, onboard sensors determined what face of the tetrahedron (a solid bounded by four equal equilateral triangles) it was resting on and then the lander opened up so the rover was upright. Fortunately, the lander came to rest on the proper face of the tetrahedron with the rover already in the upright position, but if it hadn’t, the lander would have opened the face of the tetrahedron it was sitting on first to properly position the rover.

Once the lander had completely opened up, the Spirit rover deployed its solar arrays, the mast the cameras sit on and the antennas. This is the configuration Spirit is currently in, but it must still unfold its two front and two back wheels and then drive off the lander before it can fully start its mission. Just a few final steps in the long, involved process of exploring Mars are left, and I anxiously await the pictures and data Spirit will return when it starts its exploration.

Update: The animation I am describing can be seen here.

Update 2: Wretchard from Belmont Club comments –

Is the expense worth it? Pictures of soil disturbed by the retraction of airbags show a surface that resembles mud, in a place where the presence of water is wildly improbable. Something is holding the topsoil together like play-doh and there is no obvious terrestrial explanation for it. Mars is truly another planet. And if it is anyone's it is man's. The story of Mars, its forgotten geological history and its very fate are now part of the human story. Only we will weep for it. Only we will laugh on it. Only we will remember it.

Mars In Color

The first color pictures from the Spirit rover can be found here.

Comparing Beagle 2 and Spirit

Chris Hall from Spacecraft explores the differences between Beagle 2 and Spirit in his post A Tale of Two Spacecraft. Chris states that -

The Beagle was developed as a small spacecraft, with a correspondingly small budget. While its failure (if it turns out to be so) will be unfortunate, such failures are to be expected with high-risk projects such as this one.

I would add that for a spacecraft to successfully land on Mars, everything must go right. If even a small part like a sep-nut fails, the whole mission fails. Until we can afford to launch spacecraft to Mars that have backup systems, sending landers to Mars will continue to be very high risk projects.

Spirt Coverage

If you are looking for an excellent source of information on the Mars Spirit rover, head over to Joost Schuur's Martian Soil.

Pictures Of Mars

I am watching the NASA channel which is showing the first images Spirit is broadcasting from Mars. The images are amazing, and they bring back memories of the time when I was a young child and stayed up late to watch the very first images ever taken from of the surface of Mars that were broadcast from the Viking lander.

The Spirit images are much better than Viking images from so long ago, and this mission promises to return even more spectacular data in the future.