NASA: All is Well on Boeing 747SP Open-door Front

That's an awfully big hole and there are a couple of reasons that the test pilots and engineers are being so cautious about the testing.

First off the rear section of the fuselage is often subjected to considerable torsional forces that try and twist the rear of the aircraft off whenever any sort of rudder input is applied. Normally the fuselage is a closed pipe or tube like structure which can easily transfer these forces without major distortion. However, I you have ever tried to twist a pipe that has any sort of crack, hole or opening you will quickly find that such a defect reduces its ability to resist torsional distortion to nearly zero. This is compounded on the 747SP as the reduced length of the aft fuselage means that a larger tail is needed in order to keep the aircraft under control and that means a greater force and more twisting.

I don't know what modifications they have made but it would certainly have to have been considerable as cutting a thumping great hole in the aft fuselage would just about certainly cause the tail to fall off the first time you use even slightest amount of rudder.

Secondly is the rigidity of the fuselage. The pressurized cabin of modern jet airliners is an integral part of their structural strength as it make the entire structure more rigid and far less prone to distortion. Whenever the cabin is depressurized not only do the pilots have to decent below 10,000 feet to prevent the passengers and crew from asphyxiating but they also have to slow the aircraft down and not use full control surface deflections.

The 747 is one of the strongest aircraft ever built and can really take some abuse so that would probably make it a prime candidate for such a drastic modification. I guess NASA knows what it's doing but I would be very weary until the modifications have shown that the tail isn't going to fall off the first time you try and land with a cross wind.

Hi Masu,

the telescope compartment is separates from the cabin's operator section.

I am sure they tried everything possible to calculate and assess risks.

This is a long term international project.

The most difficult part will be the adaptive optics to correct a highly turbulent situation in the optical path.

I came across one of the involved engineers in 2003 as I wanted to build my own private adaptive (visible) optical system with a 7x7 corrector matrix.

I had to drop this by lack of funding and lack of a technically top-of-the-notch technician.

The very best to you for this New Year 2010!

RHABE

"The very best to you for this New Year 2010!"

And to you.

"I wanted to build my own private adaptive (visible) optical system with a 7x7 corrector matrix."

That sounds like an interesting but very costly and complex project.

Personally I'm limited to a 125mm Maksutov-Cassegrain that has phenomenally good optical performance. However the drive mechanism and controller for it have been giving me nothing but grief. The first telescope couldn't keep track of where it was pointed and after three returns to the Australian dealer they replace it with a new telescope. The second telescope had problems with the internal wiring and also wasn't that good at tracking objects and sort of worked but not as well as I would have expected. After about two years of marginal operation I was so pissed off with it that I decided to pull the drive system apart to see if I could do better. I found a whole stack of problem but ultimately managed to get the drive mechanism working even better than it's specifications. Then the bloody controller decided to chuck a major wobbly and categorically refused to communicate with the telescope or outside world. It would turn on and you could get it to say it was ready to communicate with the telescope and my computer but that was as far as it would go. After yet another return to the dealer and them trying for several weeks to get the thing working they decided it was buggered and that the only solution was to purchase a new controller at a cost of AU$200.⁰⁰ . That was fair enough, but there weren't any spare ones in Australia so I had to wait for another month to get hold of a new one from the US. Then the weather went bad and for nearly two months there wasn't a lengthy enough stretch of clear sky to get the darn thing set up.

Finally last Saturday the conditions were perfect, but when I got everything set up the bloody drive mechanism decided to start sticking. While the telescope appears to at least be keeping track of where it's pointed it judders across the sky in chunks about half the width of its field of view. That's bloody annoying when it comes to viewing but makes trying to generate any sort of image totally impossible.

Arrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr.

It looks like it's back to the workshop and yet another rebuild.

Of course since I found that the telescope isn't working properly the weather has been perfect for viewing. I can just about guarantee that the moment I get the bloody thing working properly again it will be 100% cloud cover every night for six months.

PS: Until just on three years ago Australia had been experiencing the worst drought on record. However, the drought broke in a fairly impressive way just before Christmas 2006 and I'll give you one guess as to which piece of scientific equipment was delivered the previous day.

Hi MAsu,

I too have a rugged (but only 100mm) Maksutov-Cassegrain but mounted only on a hand driven mounting from Zeiss-Jena.

I had bought also (second hand so not this expensive) a Takahashi Mevlon 180mm on a

Losmandy G11 mounting, not yet tried - lack of time and good place and only 3 to 10 good nights here.

The adaptive optics system was intended to have a 150mm diameter, 0.5mm thick vibrating mirror with 7 x 7 exciters by self-made voice-coil actuators. The coils made from aluminum wire, isolated by oxide for allowed high temperature and best heat transfer, cooled by forced air through the channels between individual coils. The coils mounted to the mirror via a flexible joint to allow for slopes in the mirror shape.

The permanent magnet assemblies calculated for 1 Tesla flux density. The 1 to 3 V amplifiers to be bought from an external supplier (beautiful cameras for low level low noise astronomical use). This would give at near 100W permanent power the ability to drive the mirror up to 500 Hz at 3 µm amplitude. (Theory).

The Hartmann-Shack wavefront-sensor to be made by compression into a polycarbonate material by either assembled glass- or sapphire-balls or may be steel or Si3N4-ball-bearing-balls.

The individual images of the guide star resolved on a good Peltier cooled camera and the movements of the individual spots recalculated by an FPGA (2KHz cycle time) to do the correlation calculations. (and then feed the amplifiers to drive the voice-coil-motors).

All plans collapsed as the technician decided to move and no money nor qualification to do his work externally.

Such is life!

After retirement I am now beginning with experiments on interferometry to establish some high precision machines with integrated interferometers to ensure straightness, right angles and "no" spindle movement except ideal rotation.

Enjoy your scope!

RHABE

I finally managed to do some more experimenting with my telescope. This round was intended to test out how it coped with polar alignment. It worked fairly well, but the position of the eyepiece coupled with my limited mobility made viewing a considerable chunk of the southern sky impossible. There was also a problem with the telescope body fouling on the mount that prevented the telescope from being pointed less than around 20° above the northern horizon.

Unfortunately it looks like I will have no option but to use a alt‑azimuth mounting, which is somewhat disappointing as I intended to use the imaging system to take lots of pictures I can show to my fellow CR4 denizens.

It's not all bad news though. I just downloaded the latest version of the software for the imaging system and it now contains a drizzle function.

For the uninitiated drizzle was developed for the Hubble space telescope and was used to increase its field of view. Basically it works by taking a specific guide star and then moving the telescope up‑down‑right‑left while keeping the guide star within the field of view and taking multiple images. The images are then stacked one on top of each other so the guide star is always in the same position. This ultimately gives you a field of view that can be up to a little less than twice the telescopes field of view. It can also be used to overcome imperfections in the telescopes tracking mechanism that causes the image to move about within the field of view.

However, if you utilize two guide stars the system can also get rid of the rotation that the alt‑azimuth system produces as it tracks across the sky.

Utilizing an alt‑azimuth mount with drizzle to de‑rotate the image isn't the ideal solution. But it will allow me to spend more time observing and less messing about with setting up a polar alignment and performing the acrobatic techniques need to look through the eyepiece.

Really, you guys know more than I do, but I think even you have some reservations about this.

To my eyes it just looks stupid.

I understand from having even a fairly low cost digital camera that it can smooth out vibrations and jostles, gut gee whiz, shooting pictures, or focusing on anything out the side door of a plane is asking a lot.

Maybe somebody knows things I don't.

But gee whiz, holy cow, why not just go up in a C 130, open the back ramp door and let a ball turret out on a couple of cables if there is some real advantage to doing this sort of thing?

Maybe I'm just too stupid to see the wisdom of this sort of design.

"But gee whiz, holy cow, why not just go up in a C 130, open the back ramp door and let a ball turret out on a couple of cables if there is some real advantage to doing this sort of thing?"

There are a couple of problems with doing it this way:

1. The turbulence created by the tail planes and fin would create a hell of a lot of distortion that would be extremely if not impossible to correct with the current technology.
2. Second and most importantly is the centre of gravity. Moving things forward and backward which in this case is a 20,000 kg telescope, alters the aircraft's centre of gravity. Normally this is accommodated by altering the deflection of the elevator control surfaces on the horizontal portion of the tail assembly. Before any aircraft takes off the aircrafts centre of gravity needs to be calculated and then checked against the aircrafts performance tables to make sure that not only is the centre of gravity within a given envelope but will remain there for the entire flight. It's a lever type problem and the further aft you move something the worse the problem gets and the more control input needs to be added. Sticking 20 tons of telescope mounted at the end of a pole out beyond the rear of the aircraft is something I doubt any aircraft could cope with.

"I understand from having even a fairly low cost digital camera that it can smooth out vibrations and jostles, gut gee whiz, shooting pictures, or focusing on anything out the side door of a plane is asking a lot."

Your 100% correct and getting telescopes to work on aircraft is definitely a complex engineering challenge. Nevertheless it has been done before and the Kuiper Airborne Observatory (see image on right) was in service for over 20 years until it's retirement in 1995. So it's been done before and were not talking about a totally untried and untested concept.

The stabilizing is primarily achieved with a series of gyroscopes and there are two ways to go about it.

1. First off is the big hammer approach. What you do is attach a series of really big gyroscopes to the system and spin them as fast as possible without them flying apart. Take a platform and stick three gyroscopes on it so they are all perpendicular to each other and you have a platform that is almost impossible to rotate in any direction.
2. Second is the finesse approach. Rather than having a really massive rotor spinning really fast you reduce the mass to zero and increase the speed to the speed of light. There are several ways to do this but one involves taking a couple of kilometres of fibre optic cable and winding it around a bobbin or spool. You then take a LASER beam and split it into identical beams that are then sent down the fibre optic cable in opposite directions. After the two beams have travelled down the cable they are then recombined and some interesting things take place. Now, if the spool is not rotated nothing happens, however, rotate it around its axis and the Doppler effect causes the beans to change frequency slightly dependant on which way and how fast the spool is being rotated. This then causes interference patterns when the two beams are added any by measuring these interference patterns you can calculate which way and how fast the spool is being rotated. This signal is then sent into a control system which uses some sort of ram or servo mechanism to keep the telescope pointed at the same piece of sky.

I'm not 100% certain which method is being used here and they could even be using both but as mass is always a negative when it comes to aircraft I would hazard to guess that the LASER gyroscope would be the preferred method.

I hope that has helped and you can learn more about the Kuiper Airborne Observatory by following the link.

I took the tour. Thanks.

Bhankiii, Rhabe, and Chrisg288 might be interested.

Maybe ball turret doesn't quite work for CG with that plane, which I don't know well at all.

True enough I do not know of gliders towed as during WWII, suitable for the work.

Thanks

"True enough I do not know of gliders towed as during WWII, suitable for the work."

That's an interesting concept and it might just work, although I don't know of any existing glider that would be large enough to lift the 20 ton telescope and all its equipment.

It's also pretty difficult to keep a glider positioned behind a tow aircraft and can be potentially extremely dangerous.

First off you have to keep the glider out of the tow planes slipstream as this can cause the glider to be severely buffeted which not only is uncomfortable but has the potential to damage the glider. You can pass through it but you should try and limit the amount of time spent in the turbulence. The turbulence would also distort the image so to get a clear field of view you would have to keep the glider above the tow planes slipstream, which brings me to the second and potentially more dangerous problem.

If you are flying in what is called the high tow position where the glider is above the turbulence the tow plane creates you have to be extremely careful not to get too high. If you do and it's really east to, then the glider can pull the tow plane's tail up. This then pushes the nose down and increases the speed of the tow plane which then increases the tension in the tow rope pulling the tow planes tail even higher and bingo you have an out of control situation. It only takes a couple of seconds and before you can do anything about it both the tow plane and glider end up in a vertical dive at speeds in excess of V_ne¹. This also places so much tension on the tow rope that it can jam the release mechanism in both the glider and tug. The only chance you have is to try and get the weak link in the tow rope to break, but even then pulling out of a V_ne plus dive may not be possible before you run out of sky.

Originally when a glider was preparing to release the tow rope the pilot would move it through the tug's turbulence until it was in high tow. It would then release the cable with the glider performing a 90° right hand turn and the tow plane a 90° left turn. However, there were several fatal accidents in Australia that were attributed to the glider getting too high so the use of high tow was dropped from the release procedure. Glider pilots are still taught how to perform a high tow, but it's only used when there is no other option.

Anyway, it would be technically possible to use a glider to fly the telescope, but after closer inspection I'm not sure it would be practical.

Note 1: The abbreviation V_ne stands for Velocity never to be exceeded and is basically the highest speed that the aircraft can be flown. Unlike with cars the faster an aircraft moves the more effective the control surfaces become and you have to limit the amount of control input you use otherwise you can overstress the airframe. When you reach V_ne then the controls become so effective that any movement of them will cause so much force that it will damage the airframe so it's really important that you never fly faster than V_ne.

Note 2: While I am not current I have several hundred hours as pilot in command of both gliders and single engine light aircraft. However, even though I first learnt to fly in powered aircraft I enjoy and have much more fun flying gliders than a powered aircraft.

Hi Masu,

big gyroscopes for stabilisation: I was in a group that tried this for a horizontal-plane

stabilisation.

The "payload" is suspended in a cardanic system - minimum 2 axis but up to 5 rings nested inside each other, depending on the movement of the carrier.

The aircraft, ship, tank is moving (the linear accelerations are transmitted to the payload) and the angular movements of the craft shall not be transmitted to the payload.

So the bearings in the gimbal rings shall be as good as possible - but never good enough. and many requirements of static and dynamic balancing.

The "big hammer -brute force" approach is mounting these high mass high speed flywheels (angular momentum/mass to be maximised) on the payload - sometimes named "platform", so that the gyroscopic torques stabilise against the friction of the bearings that transmit torques to the platform because the craft is angularly moving around. This "big hammer" has some severe drawbacks:

1. Mass of the flywheels is dictating big bearings in the gimbals so high friction and high disturbing torques.

2. Unbalance (residual after good balancing and change of unbalance as function of motor temperature) is resulting in 1n (spin-speed) radial and axial vibrations. These are not tolerable in any high precision (arcsec) application.

The "finesse" approach is using minimum 3 independent single axis gyros (very often until now mechanical ones, so in Hubble Space Telescope!), often up to six in a package to get redundancy.

3 axis would provide stabilisation against roll pitch and yaw.

The outputs of the gyros are coupled via amplifiers for direct and decoupling (cross coupling) the axis to torque-motors of the 3 (?) stabilised axis. If two-axis gyros are used (dynamically tuned free rotor gyros) then one gyro has two measurement axis. (In the space probes now at the shockwave-front between solar wind and galactic plasma the gyros from Teledyne are this type, running at 100Hz since 1976!)

The failure of so many of the Hubble gyros is not a problem with the ball-bearings but a problem of the current feed-through to supply the floated gyros with energy. The "inert" suspension fluid is not so inert and some current will help to corrode the tiny wires if there is a minor amount of whatever conducting fluid dissolved in the inert fluid. Some change to conducting by producing acid. Voltage and current helps by this chemical action.

More modern Laser gyros suffer from ultrahigh purity requirements of the laser-gas and from many many other error sources. Modern fiber-gyros suffer from lack of stability (drift). More modern hemispherical oscillator gyros (wineglass in radial vibrations) may prove to be reliable and rugged.

I will have a look if I can ask or find some details of the stabilisation scheme.

RHABE