I work at the Norwegian Defence Research Establishment (FFI) at Kjeller outside Oslo, where I am a Senior scientist (sounds cool, eh?). I am, of course, a civilian (not military) employee, as is nearly everybody else there.
So, how did I end up at the FFI?
It all began when I started my compulsory military service in October 1994. At first I served in the coast artillery where I was stationed at Bolærne Fort, which is located on an island in the Oslo fjord, not so far from Tønsberg. Fortunately, after three months I was transferred to the Norwegian Defence Research Establishment (FFI) at Kjeller, right outside Oslo.
Although, I hadn't exactly been desperately unhappy at Bolærne Fort, coming to FFI was really like heaven to me! In addition to actually being able to do some interesting work, the accomodation was much better. All the soldiers who didn't have their own apartment stayed in a place called Perminalen, right next to Stortinget (the parliament) in downtown Oslo. It is no great overstatement to say that this was a much more convenient location to live in, than on an island outside Tønsberg without hardly any transportation into town!
During the next nine months at FFI, I mostly worked on how to describe a battle/war by mathematical equations. I ended up developing a small computer program which would solve the equations of a specific battle, and predict the outcome of the war. Obviously, it wasn't possible to consider every detail of a war with this very simple approach, but the equations still produced some results that I thought were quite interesting.
As my military service was coming to an end in September 1995, I was asked to stay on at FFI for six months and work for real. Eventually I was employed permanently.
I now began working on a different project whose title was "Protection against modern weapons". The background was that in the later years much more precise weapons had been developed. As had been seen especially during the Gulf war, it was now possible for projectiles to actually hit the specific targets they were aimed at. This was, of course, a good thing if you were the guys attacking, but it was bad news for those guys who wanted to defend themselves.
So, now the question was whether Norway's military installations were sufficiently protected against these new weapons. But, in order to say anything worthwhile about this (i.e. something based on scientific facts and not just speculation), it was necessary to know the capabilities of the new weapons against specific structures. This was where I (and the other guys working on this project) came into the picture, as we were supposed to work this out.
In earlier times, bombs and projectiles usually didn't hit anywhere near where they were supposed to hit. As a consequence, the strategy had been to fill them up with as much explosives as possible, in order to create maximum damage to the surrounding area and hopefully damage something important. Therefore a lot of scientific work had previously been done in studying the pressure blast waves (both in air and ground) generated from an explosion. In contrast, very little work had been on on the penetration of projectiles into targets.
Now, things were different as a projectile could be directed right at the most important structure (say the headquarter). Thus, it was no longer enough for the structure to be protected against a blast wave from a detonation far away, if a projectile could hit and penetrate far into the structure before detonating.
To ensure that a specific structure was properly protected, we had to find a way of predicting accurately how far a specific projectile could penetrate into the structure, as well as how much other damage it could cause. This would obviously depend on the (material) properties of both the projectile and the target. On discovering the relevant physical relations, we would in principle be able to find the optimal solution for protecting a specific structure.
My part of the project was mostly about studying the basic physics involved in a penetration process, but there also were other parts dealing with protection of specific military installations, as well as more "operational" considerations. For example: What would be the better option? Having stationary units that are very well protected (maybe by massive concrete walls) but easy to locate and attack, or to have mobile units that are not very well protected, but which are difficult to locate and attack because they can move around so easily?
Since I didn't know very much about the physics of penetration, a lot of the time in this project was spent studying what had already been done by other people. Although, as mentioned earlier, penetration is a small field, some interesting work had still been carried out.
During the project it was decided that the problem of penetration was an issue that we should pay particular attention to at FFI. This resulted in a new project called "Penetration into concrete and rock" being started in the beginning of 1999. The new project should only deal with penetration, leaving the "operational" stuff to the old project.
Some of you might be curious about what exactly it is that I am doing during a normal workday?
A great thing about working at FFI is that projects usually run over a period of about three years, which gives us time to go really deep into the various problems. There aren't too many deadlines to worry about (except the end of the project, of course) so if we come up with a great idea one day, we are able to pursue it and see where it leads us, without being afraid of having wasted time if the idea ultimately turned out to be a dead end.
In our project,we try to approach the problem from every possible angle. I spend some time working only with "pen and paper", trying to develop simple mathematical models to explain the various phenomena seen in a penetration process. It is also important to read various journals to keep up with the scientific literature and learn about what other people working in the field are doing.
A lot of the time is, however, spent in front of the computer. We're using computers for a variety of purposes. Most important is perhaps their ability to perform millions of computations in a short period of time. Amongst other things, this enables us to create socalled "simulations" of a penetration process, which is a very useful way of learning about what's going on.
So far I've only been talking about theory. However, the purpose of a theory is to predict what's going to happen in an experiment. If the theory doesn't agree with the experiment, something must be wrong with the theory! Therefore we also do penetration experiments, both to test the theories we've been working on and to obtain further information about what goes on during a penetration process. In such an experiment we typically fire various projectiles at targets and tries to record as much information as possible. (Doing this in a sensible way is actually very difficult, so setting up the experiments requires a lot of planning. There are all sorts of things that can (and will!) go wrong in the real world, which we don't have to worry about in simulations or mathematical models!)
Finally, after all the work has been done and we have obtained some new results, we need to let other people know about it. Therefore, quite a lot of time is also spent on documenting our results in various scientific reports, papers and articles. Many people don't like this part of the job, but personally I very much enjoy the challenge of presenting complicated things in a simple way.
That's a basic outline of what work on our project is like. Fortunately, we are quite free in choosing what to do on a normal workday. Some days I might be working only with pencil and paper on some mathematical models, or reading new articles. Other days I may use the computer to run simulations, or write some program to perform other kinds of calculations, or maybe document something I've been working on. At times when we're doing experiments I might also be involved with them.
Then again, some days I do a little bit of everything. It all depends on what I feel like doing at the moment, how inspired I am. I think this is a very effective way of working towards the final goal of the project.
Another great thing about FFI is that in addition to working on specific problems, we also spend some time updating our general skills. For instance we are currently getting together during working hours to study "Game theory", which is a lot of fun. All in all, the environment and atmosphere at FFI makes it a wonderful place to work, so I consider myself very lucky to have ended up here!
When I tell people about where I work, I often find that they expect me to have knowledge of all kinds of weapons, which I don't really have. Although, I have learned quite a bit about specific weapons, military tactics etc. during my time at FFI, this is certainly not my speciality. What I find most interesting is the physics. A nice thing, though, is that I can always hide my ignorance and cut any conversation short by using phrases such as "I'm sorry, but this is classified information" or "Since this is a matter of national security, I'm not allowed to go into details" :-)
A thing which did surprise me somewhat is the lack of secrecy in the field of penetration mechanics. Intuitively I would have expected quite a lot of the work to be classified, but this is not the case at all. In fact, everything is shared openly with the whole world at international conferences. I have only very rarely been involved with classified documents. I am pleased with this because then I don't have to bother locking everything up every time I leave the office.
I think that's a basic outline of what it's like to work at FFI. For those readers who are especially interested, below I'll go into more details about our theoretical work:
As just mentioned, we approach the problem in several different ways. One method is to apply the fundamental physical theory to a penetration process. On doing so, we quickly run into mathematical (differential) equations that are very difficult, or rather impossible, to solve.
Then we have basically two ways of proceeding. One option is to use our experience and physical intuition to make certain simplifications to the problem. Mathematically this means that we leave out some parts of the equations that are not expected to be very important for the final results. Typically, this involved putting very strict limits on how the materials involved are allowed to behave (as well as some other mathematical simplifications), so that their behaviour only vaguely approximates their actual behaviour.
If we're lucky, this approach will enable us to solve the mathematical equations, but the price we pay is that the solution is only valid in special circumstances. When doing different approximations we end up with different "mathematical models" of a penetration process. The cool thing about this approach is that you get immediate insight into what goes on physically. It's not just a black box that produces results, but you can immediately see what is going on and which physical factors are the most important for the process (provided that the approximations we have made are relevant).
Another option, instead of making these approximations, is to let a computer handle the complete problem. This is called a (numerical) simulation. A computer can give you lots of nice and colourful pictures representing physical variables during the penetration process, and these are bound to impressive anyone. However, it is important to look behind the nice pictures and try to understand what is really going on inside the computer, which in this case just acts like a "black box". If you set up the problem incorrectly and feed the computer with nonsense, the output probably won't make much sense no matter how pretty it looks.
Typically we have to tell the computer how the involved materials behave, i.e. we have to create a mathematical model to describe the material and feed it to the computer. Since various materials behave in an extremely complicated fashion, this is easier said than done. The computers of today are not powerful enough to take all the aspects of material behavior into account. Another problem is that for certain materials nobody really knows exactly how they behave in all circumstances. You see, some materials behave differently depending on what you do with them, how you do it, in which order things are done, if you do more than one thing at the time, and whatnot. Since there are thousands of things one could ultimately do to a material, it is difficult to predict in advance how it's going to behave in every situation.
So what we do is create analytical penetration models, run numerical simulations of penetration processes, try to work out how various materials behave, both analytically and by testing them experimentally. We then feed the results into the computer and our analytical models, do some actual penetration experiments, and finally compare the results. If everything agrees, we're very sad because then we'd be out of work....no, seriously, we're very happy! If nothing agrees, we have to work out the reason, which can be very interesting, so we're happy as well!