ThunderStruck X-2 Speed Profile

Balloon Flight with ThunderStruckCalculating ThunderStruck X-2 Speed

Recently we spoke about the spreadsheet that we have created to calculate the speed to be achieved by the ThunderStruck X-2 craft. We took into account a range of figures and that gave us some big design changed to ensure that we could meet the Mach 1.5 speed that we wanted for the experiments on board. Those changes took into account the drag of the vehicle, the angle of the nose cone, the size of the fuselage, gravity at altitude, air density and more. For the sake of the initial calculations we did not bother with the drag of hitting Mach 1 as we believed the air density to be so thin that it would not stop us achieving the speed that we needed.

I am pleased to say that team member Todd Hampson has now incorporated the transonic factors and ongoing supersonic factors into the spreadsheet and I am pleased to say that we were right. There is very little variation in our calculated speed if dropped from 45km altitude. Lower altitudes certainly had issues, but not if dropped from 40km and above. Although we are aiming for 45km using hydrogen, it is possible conditions like potential grass fires my limit us to Hydrogen

The addition factors that we have added to the spreadsheet are:

  • Base Drag
  • Area Rule
  • Transonic Wave Drag
  • Supersonic Wave Drag
  • Friction Drag

Remember that this spreadsheet is designed to measure an aircraft in a vertical dive into the ground. You can see it slow with thick air to very low speeds. None the less we intend to transition to horizontal flight at below 10km altitude, to the remainder of the graph becomes meaningless at that point.

One of the outputs of the spreadsheet is a set of figures. Below are the figures for 45km altitude release and below that for 40km altitude release. Both heights break Mach 1 sea level equivalent.

 45km stats

Above: Figures showing the results of a 45km release.

Below:  Figures showing the results of a 40km release.

40km stats

Mach 1 at Sea Level and Mach 1 at Altitude

Simply put, altitude does not really change the speed of sound. Temperature does. It is the biggest factor. Because the speed of sound is lower at altitude where the temperature can be as low as -60C, many people feel as if we are cheating if we only break Mach 1 at the altitude that we are traversing. They want to see the speed of sound broken as if we were doing it at sea level. We have provided those calculations here. The chart automatically compensates for the increased effects of the speed of sound at a given altitude by assuming a standard set of temperatures at those altitudes. These vary by time of year and region. We will publish the set of tables that we have used for this in a future post for reference.

For our purposes, the figures used will be accurate enough for the calculations. Why do we know this? Because they do not vary much at the altitudes that we are breaking the sound barrier. The coldest air in our flight will be in the jet stream and is well below 20km altitude.

We intend to give others on-line access to our spreadsheet in the near future, when we are assured all the bugs are ironed out. At this time the spreadsheet looks stable and accurate.

Below is our Velocity Profile showing Max speed in Mach figures. Remember the speed of sound changes with altitude and this is adjusting the Mach figures for the air temperature at each altitude point. The bump near the 15km is the point where the craft decelerates going through the speed of sound. Given its proximity to the ground and the density of air, it is very possible that we will hear a sonic “boom” from this event.

 45km Velocity Profile

The graph below is the Acceleration Profile in Gee Force. The bump at 15km is showing the additional drag going through the sound barrier to subsonic speeds. there is a similar bump centred at 39.5km, but the air is so thin, it is extremely attenuated and not visible and thus it has little effect.

45km Acceleration Profile

Below is a graph showing the Velocity vs Time for the first 125 seconds. After this time, the aircraft will level out.

45km velocity vs time - 125 seconds
From the top graph – Velocity Profile – if Thunderstruck X-2 continued its dive to the ground, it would hit at Mach 0.27 or 320kph . This is a lot slower than the Mach 1.5 it achieved at 39.5km. Lets hope that the landing will be a lot smoother!

Trimming ThunderStruck for Speed

X2 shadow Trimming ThunderStruck Needs Extreme Knowledge

by Robert Brand

This post is very technical. I will try and make it a little easier to understand. I will not go into very deep into the various aspects that slow the craft, nor will I get into every aspect, just the major aspects that will cause us issues.

Designing a supersonic aircraft needs knowledge of supersonic aspects of airflow and pressure/shock waves. In a previous post we looked at the basic limiting factors and those important to getting us past Mach 1. This post will look at other factors that will cause us to make small changes to ThunderStruck to ensure we reach the maximum speed possible and get as close to Mach 1.5 as possible. We previously discussed the following:

  • Varying gravity due to altitude
  • The angle of the nose cone
  • The width of the fuselage
  • Altitude
  • Vehicle mass
  • Wing and Vehicle Drag Coefficient
  • Reference Area of the object in the direction of motion

In this post we will now look at other aspects of the design that will slow the crafts acceleration during its flight:

  • Base Drag
  • Area Rule
  • Transonic Wave Drag
  • Supersonic Wave Drag
  • Friction Drag

These factors take into account compressible air flows and incompressible air flows. Look them up, but simply Transonic and supersonic flows are compressible, subsonic flows are incompressible.  They are reflected in the items above.

If you would like to look at the Maths for these issues, there is a great document from Sydney University that can be viewed on the link below:

http://web.aeromech.usyd.edu.au/AERO2705/Resources/Research/Drag_Coefficient_Prediction.pdf

Base Drag

My knowledge here comes from rockets – same as the document. A flat based rocket does not have Base drag when it is firing its engines as the air flow does not have a pressure problem when compared to having a flat rear end! Below is a snapshot of the pressure differentials at the rear of the craft. There are more and bigger pressures not shown here, but you can clearly see the problem. as a rocket flies horizontally with its engines ignited, there is no void. The moment the engine ceases ignition, these pressure waves appear – Base Drag.

X2 Base Drag Pressure snapshot

Looks like a tapered fuselage at the rear of the craft is super important to acceleration towards the ground and again as the craft decelerates due to the thickening air density. It will need to taper from half way along the main wing part to the rear and go from 300mm to 50mm– enough for a parachute to be deployed – about 50mm. Whether we add a tapered cap, taking the final taper to a point for even less drag is not important at this stage. It will look better without the cap in drawings.

This important diagram from the linked document. This shows the flight of a rocket accelerating to Mach 1.6 (Dashed blue line) and then decelerating to to low speed (the solid black line). All the various drag issues are in this typical diagram. Base drag however is the difference between the two. There is no base drag during the rocket burn and then there is base drag once the engine ceases ignition.Drag issues in Transonic and Supersonic Flight

By gently tapering the fuselage to a point, we avoid disruption the boundary layer and any turbulence. For the X2 ThunderStruck flight the fall and acceleration will also look like the deceleration. Base drag will almost be eliminated.

Area Rule

We have spoken about this in an earlier post. That is keeping the cross-sectional area of the craft constant – so thinner where there is space (area) allocated to the wings. Area ruling will be somewhat addressed by the taper to the rear as discussed above in Base Drag. It is a fairly small effect unless you were spending significant time near the speed of sound. The X-2 ThunderStruck craft will spend 15 seconds between Mach 0.9 and Mach 1.2. I believe that it will be small and this is where the area rule has the biggest effect – but still small. There will be no additional change for area rule.

Transonic and Supersonic Wave Drag

The taper of the rear of the craft will minimise Wave Drag – both Transonic and Supersonic. Some playing with Wing Design may change the Wave Drag, but we will ignore it at this stage. I am not looking to play with the design unless there is a strong case. In the diagram above the Transonic Wave Drag begins at about Mach 0.9 and Mach 1.2 and Supersonic Wave Drag continues upward from that point.

Friction Drag

Friction drag occurs at low speeds with laminar flow being disrupted and the airflow becomes turbulent. We will have stalled at that stage and thus this is of no interest since we have an aircraft. We should have landed! This is ignored.

Trimming the Design

We have determined that we need to do two things. Stop the leading part of the winglets from protruding in front of where it joins the wing and to taper the fuselage. We will provide a picture of the new design shortly. Here is a render of the current X2 design without the new modifications:

X2 - Clouds2

The X2 ThunderStruck craft will have minimal impact regarding its maximum speed. I will reveal the new graphs shortly showing the speed at any given altitude point. As the air is extremely thin at our launch altitude, the increase in drag above 35km from the items above will not likely to be affect our top speed much as previously calculated, but may increase the deceleration slightly. That is the max G force as we slow. I will publish the updated results soon.

Finally a scan of the pressure waves from front to back on the X2 craft before we trim the craft:

X2 Pressure_Cut_Raised

45km Free Fall Spreadsheet

ThunderStruck Phase 1 Version 2Fine Tuning ThunderStruck’s 45km Free Fall

By Robert Brand and Todd Hampson

This post will examine a drop from 45km. Todd has done an amazing job on this interactive Excel spreadsheet. We can change a few variable and see the change effects. It has done an amazing job in letting us see what changes produce the greatest “bang for the buck”.

The first thing was changing the diameter of the craft creates a great difference in drag. We have decided that we need to make the fuselage 300mm in diameter (down from 600) as it have a huge effect on when the craft stops accelerating. It adds Mach 0.3 to the top speed. from a 45km drop. We also noticed that if we get the craft the right dimensions  and mass, the need to get the balloon to 45km is reduced. We can still break the sound barrier with a release from 40km altitude. At 45km we get a top speed of Mach 1.54 and at 40km we get Mach 1.36. This is also with a mass of 15kg rather than 10kg as we previously had though would be sufficient. We do not want to release much lower as things change rapidly with the thicker air.

Thunderstruck Drop Test Simulator

Thunderstruck Drop Test Simulator

Above is the top part of Todd’s spreadsheet, the coloured cells to the left allow different inputs and the cells on the right are the snap summary. The model that we have made only just got us over Mach 1 with little to spare. Changing the diameter and elongating the nose (a smaller 1/2 vertex angle of the cone) made a huge difference and making the mass 15km means a huge leeway. As mentioned on TV recently, we are aiming for Mach 1.5 and now we have the maths to prove that we can reach that speed. One interesting aspect of the reaching Mach 1.5 is that the deceleration by the thickening atmosphere is about 1.3G. Barely more than standing on the ground. It is a really gentle load and it is mainly on the nose cone of the craft. The wing and tail assembly will keep the craft oriented in the denser air and we will rely on the ballast in the front of the craft to keep it nose down.  The ballast is likely to be antifreeze and we can shift it or eject it for a more stable and slower level flight.

We hope to have the fully interactive spreadsheet available on the site for those interested, but until then let’s have a snapshot of the curves that count. That is a free fall from 45km.
Thunderstruck Drop Test Simulator Max Speed

At sea level, mach 1 is about 340m/s. I say “about” because air pressure has little to do with the speed of sound. It is mainly air temperature. From the graph we should reach 530m/s and that is Mach 1.56.

Before we streamlined the craft to punch through the thickening atmosphere, the wider bodied version of ThunderStruck slowed down really fast and took some stronger G force on the nose (mainly). The version 2 craft slows almost at the same rate that it accelerates. This gives a very gentle change as can be seen below.

Thunderstruck Drop Test Simulator Deceleration

From the graph above it is clear that at 45km, as the speed increases, the air resistance has a greater effect. At that height air density is about 0.025 10-1 kg/m3 compared to air density at sea level which is about 12.25 10-1 kg/m3 (plus or minus about 5%).

This means that our air density at 45km as a percentage of air density at sea level is about 0.284% that of sea level and it increases as we go lower. The effects also increase with ThunderStruck’s speed as the drag has a greater effect with both speed and increasing density.

With the calculated drag of the craft, we find that all acceleration stops at 26 km and as we fall into denser atmosphere, we begin to slow. The graph above is calculated in metres per second per second (known as m/s/s or m/s2) and that can be directly converted in to G force. Since 1 g = 9.80665 m/s2 a simple rule of thumb conversion to remember is 10m/s/s = 1G.

Now for many the next part of this may be hard to grasp, but at free fall at 45km we have what is loosely termed 0G where, if we were in a craft also falling at the same rate, we would float inside the craft. Once we reach terminal velocity at 26km altitude ( I will ignore the lag in deceleration here), we have 1G acting on the craft. If we were inside that craft we could walk around the interior and feel the same as on the earth’s surface (again small variations in gravity, etc excluded). A skydiver that has reached terminal velocity has the air flow stabilising his speed and that air flow has a force of 1G on his body. G force real is only noticeable when there is change – ie a change in direction or acceleration or deceleration.

The “vomit comet” aircraft that simulates zero G does so by moving steadily in a straight line while accelerating towards the ground at 9.8m/s2.  If they just dipped the nose and began that arc, but stopped accelerating towards the ground we would all feel an initial 0Gs but then would be back on the floor when the rate of change returned to zero and we would be back at 1G. So, with ThunderStruck, it is the rate of change that determines G force and at 26km altitude, the G force is 1G on the overall craft, but since the greatest drag is caused by the nose, the 1G force is felt here. Other parts of the craft would be happy to continue accelerating! So at 26km, the structural form of the craft must allow the nose to hold the craft by the nose vertically – good to know, but it does not stop there.

The craft continues to slow and decelerate with the denser air and we have to slow way more. That now takes us into the realm of more than 1G. In fact at 17km we experience the greatest rate of deceleration or change and that is an additional 1.25G for a total of 2.25G on the nose of the vertical craft.  That is the base amount of structural integrity we will need in the nose assembly. If the craft weighs 15kg, then the nose assembly has to support 33.75 and then an amount that we required to ensure it is strong enough. My design had better look to supporting 50kg on the nose when the craft is stood vertically at least.

It seems that what you gain, you have to give back. The higher the speed and the longer the period in low Gs, then the the higher the Gs or the longer in negative Gs you need to complete the flight back to a complete vertical stop. I have not analyses the areas on either side of the 0m/s/s on the chart above, but I would not be surprised if they where equal. As we say here – swings and roundabouts. What you gain on the swings, you will lose on the roundabouts.

Mathematics is a wonderful tool for designers. From a few simple facts in a spreadsheet, we have calculated the speed at all points in the flight (vertical dive perspective) and also the internal forces on the craft at many points. ie the winglet tips will be a point of high drag so they will need to handle more than 1G vertically. The same with other leading edges and that also goes for surfaces affected by shock waves. All of which can be determined by design and software. You don’t have to be a maths genius, but you do need to know maths enough to ensure that you can use them in day to day work. Unless you visualise what is happening, you will have an unhealthy reliance on software for everything you do. That often denies the genius of innovation. It is also why a novice can invent something a seasoned engineer fails to see.

By manipulating the graph by fine tuning the inputs we found that our craft accelerated longer or you could say “the rate of deceleration was slower” by:

  • Making the nose cone pointier
  • Making the fuselage (and the nose cone) a a smaller diameter
  • Increasing the weight of the vehicle

In fact with the new design we have found that we can still break the sound barrier at a starting altitude of 40km. that is our plan B if the weather or winds in the atmosphere go against us. ie, we can launch early if the winds are taking our balloon out of range of our communications systems.

So what does our new design look like?

This is an early look as there are a few bits at the rear that still need adjusting.
ThunderStruck Phase 1 Version 2

We also took the opportunity to correct a few other aspects of the craft:

  • Bigger wing Area with a larger area ahead of the main wing
  • Longer spikes on the winglets (the winglets are not as high due to the smaller fuselage). This is to move the supersonic shock waves away from the control surfaces on the rear of the wing.
  • Twin rudders trailing the craft (there are some wing tabs in the drawing that need to be removed.
  • A tapered tail to stop high drag behind  the craft (we also need to remove some wing tags in the model above.
  • Tapered rudders on the bottom to stop it hitting the ground on landing (not shown)
  • Tapered rudders on the top for symmetry to ensure that it has little differential in forces on the craft to make it pull out of the dive.
Bruce Boler and Jason Brand with ThunderStruck Phase One 1/6 Model

Bruce Boler and Jason Brand with ThunderStruck Phase One 1/6 Model

So there you have the new design based on maths and simulations on a home computer. It seems that building a supersonic aircraft is child’s play as Jason (12) is jointly working on this design. This morning I asked him what G force is at work on a skydiver at terminal velocity and he confidently answered “1G”. Good one grasshopper. He then went on to clearly say that g force was related to change in acceleration (relating to a skydiver). I love it when he talks maths. He needs to know as he will be the remote control pilot for this Mach 1.5 aircraft.

Calculating Maximum Speed in Free Fall

100km accelerationFree Fall Speeds

by Robert Brand and Todd Hampson

Oddly enough, there is very little information on the web for calculating the maximum speed that a craft will fall from a specific height. It is a complex calculation requiring knowledge of the shape of a craft, the size of the craft, the amount of gravitational attraction at each height, the thickness of the atmosphere and the mass of the vehicle.

Todd Hampson has done some great work in getting the information together although he has not found a simple formula for calculating atmospheric density. He has temprarily used look-up tables and that has caused some rather “jerky” graphs. He will work on embedding a formula into the equations and removing the problematic look-up tables. None the less, this is a story of our travels and thus our problems too. Eventually it will be our triumphs too, but a bumpy chart is not a major worry to me, especially as we already know the solution. Now for the fun stuff.

Calculations, Calculations and More Calculations

Getting something “just right” the first time is near impossible and this is no different. Lots of complex data and no simple formula for air density, simply because it is not linear and non anything else. Tomorrow we will add the formula into the data and smooth out the bumps.

Today let us look at the graph that is all important, but first let’s look at an version of ThunderStruck falling from 100km. We will need to do this for Phase 2 with a different craft, but let’s look at the maths.

Todd says:
– For mass of the vehicle I used 10kg.
– For the Area of the object in direction of motion (vertically downwards I am assuming for the high speed part of the fall) I calculated the cross sectional area of the cone ie: a circle using the diameter of 600mm as per the current drawings.
– For the Drag Co-efficinet there was a URL on the VUId page that pointed to an aerospace.org page discussing different drag co-coefficients. For a 3D cone the Cd is calculated using a formula that needs a half-vertex angle. From your drawings (cone depth 450mm, cone diameter 600mm) half-vertex angle is 33.7 degrees.

100km release; max speed

max speed for an aircraft released from 100km – from a sounding rocket apogee of 100km

In the graph above, the first part of the flight was a little more difficult than I thought as lots of things are changing as it falls ie: gravity, air density, drag etc but I’ve got there now.

The first model I have done is the 100km drop test. I need to clean up the data below 18000m but the show is well and truly over by then anyway, but I will get it right so the graph is correct (I need to be more accurate with the air density below 18km).

This says a lot. Thanks Todd. This shows that tourist flights to space at just over 100km altitude at apogee will reach a top speed of Mach 3 on their return – that is about 1,050m/s. Then without any further intervention, they will slow to a fall of about 50m/s near the ground. This shows that the Virgin Galactic trick of feathering the craft is all about stability and not speed. There is nothing that will prevent the craft from reaching this speed since there is not enough air to interfere with the acceleration. The “chunky” graph below shows that clearly. Please assume that the peaks to the left in the deceleration part of the graph are correct.

Acceleration from 100km fall and then deceleration

Acceleration from 100km fall and then deceleration

Free Fall Speeds

From the above, you can see the acceleration is flat and continuous until the craft reaches an altitude of 60km and the acceleration starts to slow. It crosses the zero point of a stable speed at about 47km and then begins to decelerate quite rapidly until it reaches 33km altitude. At this point the deceleration slows down and at 20km altitude the deceleration is slowing in the thick air. You may notice that the maximum deceleration is 38m/s/s and since we accelerate at nearly 10m/s/s when we jump from a platform, simply put every 10m/s/s equates (rule of thumb) to 1G. This means that any craft headed straight down will experience a maximum G force of about 4G. Nothing too harsh. Slowing from orbit is very different and we will eventually cover this in future posts about re-entry.

The first thing to notice is that we will never reach Mach 3 from a release at around 45km. We will achieve over Mach 1. There are a few things that we will need to play with to reach the desired Mach 1.5 and we will cover that in a future post as we look at the graph for a drop from 45km and another from 35km.

3D Files and ThunderStruck Phase One

Thunderstruck Half Scale3D Files, Printers, simulations and More

by Robert brand

In the past we used wind tunnels and that was fine for subsonic flights. ThunderStruck does have access to wind tunnels at a number of establishments, but it is unlikely that we will need them. In part because supersonic wind tunnels are rare and because subsonic tests do not translate to supersonic conditions. Why wont we need them? Simply because computer modelling allows us to test most things extremely accurately without  the need for wind tunnels. The first part of the equation is to “make” a 3D model of the airframe and from that the options are extraordinary. Simulations of wind tunnels are just one option. What else can you do with the computer files of your model? well, we are not doing everything possible, but here are a few things:

  • Produced images of the ThunderStruck craft in solid form (we use Solidworks)
  • Rendered the surface to appear metallic
  • Added the ThunderStruck Logo and artwork
  • Animated the control surfaces on the craft
  • Sent it to a TV animator who will use the flight profile to simulate the mission
  • Made 3D models of the craft with a 3D printer.
  • Made a scaled nose cone for the 1/6th size model for demonstrations. Nose cones are immensely hard to create, but so easy with a 3D printer
  • And finally (so far) carried out Mach 2 flight simulations
Thunderstruck Phase One plans and view

Thunderstruck Phase One plans and view 1/2 scale

These simulations show up any problems and thus they have already resulted in small changes to the Phase One craft design. The biggest change will be a longer and more slender nose. You will see why in a moment.

One the right are the original plans from three sides. The software automatically creates the view (top right of image). The 3D files are then produced and it is often that simple. Everything flows from the files. The extension for the files is STL. A printer may break the files up to print an object in two, three or more parts. It depends on the size of object a printer can handle.  we wanted a 22cm model of thunderstruck and that was printed in three parts as it was too wide and two high. The parts were simply joined with acetate. It melts the material slightly and the pieces are then welded together without glue.

The solid image looks like this with a little bit of shadow and a plain surface. A “light source” is placed where needed to create essential shadows for the right feel and look.
Thunderstruck1

The image above has been created to appear to sit on a grey surface. Remove that surface and add a metallic texture and a background image and you get this:

Thunderstruck Phase One Craft in Flight

Thunderstruck Phase One Craft in Flight. Credit Ben Hockley (ThunderStruck) and NASA (clouds and Moon)

The video below shows an animation for the control surfaces. Nothing much to see other than we are working on getting the smaller bits right for the big animations. You can also see our logo on the side so this is a two in one demonstration

Below is a rough picture of the printed 22cm model and the Nose Cone needed for our TV interview on Wednesday with Channel 7 (The Seven Network, Australia). It will be painted. If you look closely you will see the nose cone join and similarly you will see the join on the 3D model.

22cm ThunderStruck Phase One model and the nose cone on the right

22cm ThunderStruck Phase One model and the nose cone on the right

Below is the plan for the nosecone and it is simple to reproduce. Notice the curved area near the base of the nose cone. This is to ease the airflow over the surface and prevent the delamination of the airflow.

Phase One Thunderstruck 1.6 nosecone

Finally we can do simulations. I will explain what you are looking at below in the next post but wow this stuff is impressive. This si not the top end software, but just a basic system and it is more than adequate for our needs:

Airflow_temperature

Most of the work on this page has been provided by Team Member Ben Hockley of Brisbane. I am grateful that we have a person with his skills in the team.

 

Aerodynamics of Supersonic Craft

Supersonic Glider-spacecraftSupersonic Shock Waves

by Robert Brand

Phase 1 Test Craft

As you all know by now, Jason, my 12 year old son, will attempt to break the sound barrier mid year – he will be 13 by then. This is the first test of a high drag system that should limit our airspeed at supersonic speeds. It is the Phase 1 testing that we talk about in our documents. We need to go fast for this experiment, but returning from space we need to slow down. Our transonic tests will be with a very different looking craft than our future spacecraft.

To the right, you will see where Jason’s design started. As you may have noticed, the ThunderStruck’s current design looks nothing like the image to the right. At each stage he has had to modify the craft to achieve the goals of going supersonic and get up to around 2,000kph ThunderStruck Design and 1-2 size measurementswith full stability. Only then will we deploy the experiment and hopefully slow the craft dramatically. As you are aware, it now looks more like the craft to the left. These are massive differences and we will explore the choices he made in another post. Right now we are looking at how we get rid of the major shock waves and also how our Phase 2 test aircraft may look. The main differences in the 2 craft are:

  • No supersonic Spike
  • No central tail
  • Winglets above and below with a ganged rudder
  • Delta wings that have two angles of protrusion from the fuselage
  • Elevators and ailerons are at the rear of the wing
  • The wing extends to the rear of the craft

Phase 2 Test Craft

In Phase 2 the craft that we design will need to travel straight up into space on a sounding rocket. We will separate from the rocket and continue our climb (momentum) to apogee (top of the flight) and then fall back to earth. Apogee may be as high as 200km. The air is so thin that we can conduct what we call weightlessness experiments for several minutes. Once the air starts to create drag, the experiments will end as the craft will slow. At this time we do not want the craft to accelerate further, but it will. Unless we feather the wings(like Virgin Galactic’s Spaceship 2) or create massive drag in some other way (our Phase 1 experiment) we will go too fast for our craft’s well-being. We need to go slow as possible. We do not want to have to slow with unusual braking as this may de-stabilise the small craft.

Hyabusa reentry sequencWe could use a capsule, an ablative heat shield and a parachute like JAXA’s Hyabusa, but we are creating a winged vehicle, although the capsule will always be another option. I guess the cat’s is out of the bag. We will have a couple of configurations possible with capsule or winged reentry as an option. The ThundrStruck craft will be a modular design in the style and electronics. The picture to the right is the landing sequence for JAXA’s Hyabusa that landed in the centre of Australia. It is not complicated, but you do have to know what you are doing and the downside is that it lands where ever the winds take the parachute.

I want to fix that problem for those customers that need a precision landing or effectively or a smooth landing. I would love to be able to direct the returning spacecraft to a point on the map that allows us to land it without having to recover it from an unknown place in the desert.

Supersonic Aircraft SpikeThe picture at the top of page is where we started. I expect that the spike will not be on the spacecraft and it is also now unlikely to be on the transonic test vehicle, but it is important to understand why we see them on supersonic craft. Sometimes a very long sharp nose can also produce the desired effect.

The picture at right is a NASA test vehicle with a spike. There are many supersonic aircraft that either have a spike of a very sharp nose well ahead of the wings. Why? We discuss this after the following paragraph.

Returning from space the spike would be a liability in the heat of reentry. It will also not be an asset in slowing down a craft. We only need to have the spike as an option to help lower the Resistance to breaking the sound barrier for our tests. At the time of posting, Jason has gotten rid of the spike and opted for wings tucked back behind the shock wave.

In our tests we will use gravity to accelerate the test craft to way past the speed of sound, but shock waves (pressure waves) would slow us down and limit our top speed. We would probably still break the sound barrier dropping the craft from around 40km altitude, but the quicker we transit the sound barrier the higher our top speed and the better the results from our experiment.

So What Does the Spike Do?

supersonic shockwaves in a windtunnelAs I said a sharp nose is the same as a spike and the image to the left shows the effect of the nose/spike as it moves the shock wave to a point well ahead of the main body of the craft and away from the wings. A sharp point is a very low area of shock and in the image you can see the shock waves from the wings as very low level compared to the shock from the tiny front of the aircraft. So long as the wings are tucked in behind the initial shock wave than the drag is lowered considerably. The reason that it was so hard to break the sound barrier was simple. The craft used had their wings in the high drag area caused by shock waves.

Now I may have been a bit simplistic here, but none the less, the spike is important to supersonic flight. Since we are wanting to slow down in Phase 2 tests returning from space via a sounding rocket, we can actually round the nose of the returning spacecraft and still get the supersonic shock to clear the wings

So Why Didn’t the Shuttle Need a Spike?

WPointy nose and shockwaves at mach 6.ell it did need to slow down and so you might think that a blunt nose is a good thing to create drag, but that is not the reason. Wouldn’t a sharp nose be good for takeoff, spike or no spike? Well, in some ways, yes, but the shuttle had wings that were very wide and a spike could not be placed that far forward. The resulting shock waves on takeoff and especially re-entry would be a bit problem as they would hit the wings.

Re-entry would be the biggest problem. The shock wave from a sharp nose would hit the wings and further heat the air. You would be adding thousands of degrees to the heat that it is already being generated on the leading edge of the wing – not a good idea! See the image above right. This would be a poor design for such a craft. The image shows a pointy nose model in a mach 6 airstream. You can see the shock waves hitting the wings midway along their leading edge.

So What Happens with a Blunt Nose?

The image to the right says it all. The blunt nose acts as a ram and pushes the shock wave way to the sides. This misses the wings by a long way – and the tail of course. The blunt nose does add to drag so that is another benefit to slowing down, but a minor one. It is the additional heat caused by the shock wave over the wings during re-entry  that had to be eliminated

What Else Protected the Shuttle from Shock?

Ever consider the orange main fuel tank? Where was the shuttle positioned relative to its nose. It had a point, but was really broad.

What effect did that have during launch at high speeds. The shock wave that resulted missed the shuttle entirely. It is important that the top of this tank was far enough forward to protect the shuttle. The whole design and shape of the combined modules on the launch vehicle was super critical and not just a random bunch of sizes. Minimizing shock waves means being able to both protect the vehicle and increase the payload as you have less drag.

In other words, if the main tank had needed less fuel and had been smaller, then it would still have needed to be as high to push the shock waves aside.

Each and every part of an aircraft that changes its size or sticks out causes shock. You must account for it or suffer the consequences.

The image at right clearly shows the shock wave of the jet disturbing the water. You do not have to be traveling at supersonic speeds to produce shock waves, but the faster you go, the more power is lost and the stronger the shock wave.

Aircraft Design Changes

ThunderStruck mk2ThunderStruck Design on the “Fly”

We now have 2 major changes to the ThunderStruck aircraft. The first is shown in the image to the right. Winglets. The second is a square-ish cross-section to the fuselage rather than a round fuselage – this is under consideration to aid in landing the craft.

The image to the right still shows the aircraft with a round fuselage, but it is obvious that we will miss out on some lifting ability from the body at landing by making the fuselage round. A flat surface on the underside if the craft will provide more lift at the right angle of attack.

winglet_effect_Winglets

The image right is from Wikipedia:

Line drawing of wingtip vortices behind a conventional wingtip (on the left) and a blended winglet (on the right).

This is important as it reduced the vortices behind aircraft that cause so many dangerous incidents at airports when aircraft get too close to each other. It also reduced drag and thus efficiency in aircraft.

The ThunderStruck craft will certainly use the winglets to reduce drag by reducing vorticies, but this will have no impact at supersonic speeds because we will be using symmetrical wing. Normal wings have a flat bottom and a rising leading edge and a trailing edge on the top surface.  This makes the air flow faster over the top compared to the steady flow over the bottom. This reduces the pressure on the top pushing the wings up. This would be nice for the cruising stage after the dive, but bad for holding the craft in a supersonic dive. Any air flow over any asymmetrical surfaces may produce drag or lift that could pull the craft out of a supersonic dive early. The effects could be catastrophic.

Many high altitude model aircraft dropped from high altitude balloons (usually illegally) follow a roller coaster ride due to the thin air and lift in the wings. We don’t want that so the wings will be symmetrical – no lift. We do not need them for the supersonic dive. What we also need is symmetry in the aircraft at any cross-section, vertical or horizontal. The closer to total symmetry, the more likely that ThunderStruck will reach speeds of near 2,000kph. So if we have winglets, they need to extend top and bottom.

So Why the Winglets?

Simply we need wheels. The winglets hide the wheels and any need to lower wheels for landing. We may use a retractable wheel for the front, but not the rear wheels.

The Winglets will also house twin rudders, making a dedicated rear stabiliser (top and bottom) unnecessary. The rear cross-section looks like the picture below:

ThunderStruck Cross-section

Lifting Body (at Landing).

In the cross-section above the flat surface of the lifting body is obvious. This will only be important when landing as the craft assumes a significant nose down attitude during the gliding phase. Since we have no lift from the wings, the craft needs strong elevators to redirect the airflow at the rear of the aircraft to keep control. We will stay aloft by having speed due to a high angle of attack (nose down). Large elevators will keep the aircraft flying at this high angle of attack. It will be a poor glider – but so was the space shuttle – for different reasons – more to do with the delta wing configuration. A round fuselage cross-section would not aid the lift of the craft at landing. A square fuselage will increase the drag as the surface area is greater, but it will help fly the craft at lower speed when landing. It will have little effect during the glide phase. We may add canards to the front of the craft to increase the lift at the front during low speed flight, but they will pop out after we go subsonic. Delta wing craft work well at supersonic speeds, but are poor performers at low speeds. In the picture at the top of screen, the craft does not have a supersonic spike. We will need this for the Transonic tests, but not for return from a sounding rocket or re-entry from orbit.

Below is a closer look at the Winglets. We have yet to show the square cross-section in an image, since this is still under test. It is felt that the flat surface will help drive a higher pressure under the craft (between the ground and the craft) allowing it to land at a slower speed. This is a form of “ground effect” making the need for a long runway important to drop off speed until the effect lessens and the aircraft eases to the runway. Tests may find little difference in the landing speed and thus we may revert to a cylindrical fuselage. Time for some wind tunnel testing.

ThunderStruck mk2 closeup

 

Preparing for the Flight of ThunderStruck

Weather balloon burst

What a burst weather balloon should do! Disintegrate

ThunderStruck – Backup Preparations

Jason, our 12 year old pilot is no stranger to having to prepare for the worst and it is what we do every time we send up a payload. Our last flight of a balloon into the stratosphere was a case of just that. Two failures. One on launch and the second on decent. Each problem would be enough to cause most balloon payloads to be lost, but as part of our preparations, we carried two trackers for the one flight. This was a flight in preparation for our project and we are testing. We have had to cover our payload in the video. Our apologies.

Below: An artist’s view of the ThunderStruck aircraft under a zero pressure balloon (more on that another time) at 40km altitude. You may have guessed, I am the artist….. Note that on the ThunderStruck event, we will not be using weather balloons so there will be no unexpected explosions.

Balloon Flight with ThunderStruck

Failure One

The first failure was totally invisible to us. A massive downdraft. The first that we have ever encountered. Uplift-1, our first flight, started in an updraft and it rose at an incredible rate for the first kilometre. In the video below, you can hear me make the comment that there did not appear to be the lift that we knew we had because we had used scales to measure the lift. We could not feel the downdraft pushing the balloon down 15 metres above our heads. I mistakenly thought my lack of “feel” was because of the others also holding the payload. We released the payload and balloon and then our hopes sank as the payload only lifted slowly and then sank back to the ground. We ran to catch it, but it rose again and caught on the edge of the eve of the roof of a nearby wheat silo. It stayed there for only 2 minutes, but it felt like an eternity before it released. It rose quickly as calculated, but one tracker had had its GPS unit disconnected and the other had its antenna twisted 90 degrees effectively lowering the power considerably. None the less we could still track the flight – mostly.

One tracker disabled, but still sending its ID at full power, The other effectively made to look low power. Those GoPro cameras are great. hundred of metres above the ground you can hear (faintly) people talking and a dog barking! They make great gear.

Failure Two

The weather balloons are meant to explode and disintegrate. This one did not. The entire balloon, well over 1Kg fell into the parachute and tangled itself in the chute, effectively making the mass look like more like a tangled flag than a parachute. It slowed the payload in the thick air, but the fall from its maximum height was rapid and the entire fall from 30km only took 15 minutes. This was an average speed of 120kph. Given that the payload probably hit the ground at 30 to 40kph, the initial speed was probably close to 400kph in the thin upper air.

With the tracker only giving us effectively a poor signal, the last track that we received in one of the vehicles headed to the landing site was 2 km above the ground making the landing site potentially one square kilometre.  We also fond out later that the second tracker was never going to give us a signal, because the impact had caused a battery to eject from its holder. We only had one ID every 20 seconds and no GPS location! We used a directional antenna to lead us to the payload, but it was a slow and painful task.

The video below shows the impact and the wooden spars breaking. The camera continued to record! Nothing like a good wiring system to ensure that power kept flowing from the external battery. I did not mention that we use external batteries. The GoPro’s batteries, even with the additional power pack, just do not last for the entire flight if it goes over 2.5 hours and especially if it is taking both videos and stills – The new GoPros are amazing, but need more power for High Altitude Balloon (HAB) flights.

Initially the video above shows the incredible stability of our payload at 30km altitude. The Balloon explodes at the 30 second mark and then plummets and spins at a sickening rate of a  couple of times a second with the disabled chute causing the spin.  At 1 minute 45 seconds, we cut to an altitude of about 3km and it took 3 minutes to hit the ground at 60kph. At the 4:45 mark, the payload hits and spars shatter. The camera keeps recording. By the way, the big tree lined road is the Mid Western Highway. The payload was kind enough to land in a sheep paddock beside the main road. You can’t ask for better.

The Lesson

The lesson here is that if it can go wrong, it will go wrong. Yes, we have recovered every payload that we have sent up, but good preparations both in the payload design and build is important as are the preparations for recovery on the ground. We even carry poles to remove the payload from trees. We can manage 14 metre trees. After that we will have to look at other methods.

Our preparations will be backup, backup and more backup. Redundancy rules over weight considerations where possible. Systems will be over-engineered and more care will be taken than what appears necessary. Project ThunderStruck will fly while the world watches. Delays will be unacceptable. This was UpLift-20 and again we have 100% successful recovery rate. @0 flown and 20 recovered. As our flights become more aligned to the actual shape of the ThunderStruck aircraft, speeds will dramatically increase on decent and the videos will have way more interesting stuff to show, but these lessons were there to remind us not to get complacent.

The View from 33.33Km Altitude

33.33Km and the Thin Blue Line

UpLift-19 Media and Information

Ever wonder what the view is like 1/3 the way to space. Here is our last high altitude Balloon flight to give you a look. Since it did carry sensors for Project ThunderStruck, it is a big thank you to Clintons Toyota of Campbelltown, NSW, Australia. The balloon was launched from Rankins Springs NSW and the payload weighed 1.5Kg and it was a 800 gram balloon

This is an unedited video and still video images from a GoPro3 Black edition camera of a weather balloon payload area. It climbs to 33.333Km where the balloon bursts and the payload free-falls back for recovery. It was a commercial flight fo Clintons Toyota, Campbelltown, NSW, Australia. They also sponsored a non-commercial payload for Project ThunderStruck – our first test for the Project for a supersonic glider to break Mach 1.5 (1,800kph / 1,120mph)

http://projectthunderstruck.org

The so-called Space Chicken, frame and with the parachute deployed, it reached a top speed of 400kph / 250mph. At the 12 minute 14 second mark on the video (2 hours into the flight) there is a noticeable jarring of the payload and a small pop. This is the balloon exploding. Immediately shredded balloon hits the payload as there is virtually no air to slow it. 2 seconds later, the payload tilts showing the cloud of shredded balloon About 1 minute into the free fall we reached 400kph according to the telemetry. The drag increases at lower altitudes, so the effect of the wind is worse as it descends. It then improves as the air density increases. In the seconds after release you get to glimpse the balloon shreds rocketing into the payload from the explosion and then the cloud of shredded material in the sky. About 10 seconds later there are glimpses of the blue and white parachute not doing much during the fall due to the low air resistance. The cutdown box that is placed above the parachute actually fouls the parachute slightly during the free fall before it becomes effective at slowing the payload. The fouled parachute causes spin at the faster speeds. The video finish with the payload still well above the clouds. This was UpLift-19 by Robert and Jason Brand for Clintons Toyota.

PS, notice that thin blue line in the video and the photos? That is all the atmosphere we have and that is pretty thin near the top. 72 percent of the atmosphere is below the common cruising altitude of commercial airliners (about 10,000 m or 32,800 ft)

Jason and Robert Brand setting up the cameras on UpLift-19

Jason and Robert Brand setting up the cameras on UpLift-19

 Balloon-Burst1-seconds-after-the-event-UpLift-19

Balloon-Burst1-seconds-after-the-event-UpLift-19. Those are the shreds of the balloon.

Balloon Burst3 seconds after the event Note the cloud is getting smaller as the thin air slows it faster. – UpLift-19

Balloon Burst3 seconds after the event Note the cloud is getting smaller as the thin air slows it faster. – UpLift-19

Balloon Burst4 seconds after the event - UpLift-19

Balloon Burst4 seconds after the event – UpLift-19 – yes, that is the sun

Balloon Burst5 seconds after the event - UpLift-19

Balloon Burst5 seconds after the event – UpLift-19

Balloon Burst6 with Parachute in view seconds after the event - UpLift-19

Balloon Burst6 with our blue and white Parachute in view seconds after the event – UpLift-19

Balloon Burst7-Effects of drag are clear after only 24 seconds - UpLift-19

Balloon Burst7-Effects of drag are clear after only 24 seconds – UpLift-19

Balloon Burst8 - Speed has slowed, but drag is greater in the thickening atmosphere - UpLift-19

Balloon Burst8 – Speed has slowed, but drag is greater in the thickening atmosphere – UpLift-19

Note: The images above are from the High Definition Video, not still images. The quality of our camera work has increased dramatically with some improvements to our methodology.

Stability

Creating Stability Between Supersonic Dive and Subsonic Level Flight

Here is the problem. During the supersonic dive, the weight is ideally forward to ensure that, as an airflow is felt by the aircraft, the drag of the tail keeps the craft oriented in vertical dive. That is assisted bythe drag across the aircraft and a low centre of gravity near the nose. During level flight below supersonic speeds the centre of gravity must be further back and ideally between the wings.

Jet fighter design has all sorts of tricks to alter the centre of gravity (or appear to) to make he changes needed. This can be as simple as changing wing shape or even extend more wing during lower speed flight. Some aircraft even have had swung wings. It is hard to control the major variations between the lift and drag that changes dramatically between subsonic and supersonic flight. None the less they do not need the dramatic changes in the centre of gravity that we are engineering.

ThunderStruck will be essential a poor flier as we are, at this stage, proposing symmetrical wings. The problem is that nothing is perfect and even the subtle differences between the wings can give on more lift than the other and create spin. Because of Bernoulli’s law, you might have supersonic flow on the wings, nose, or any other curvature way before you reach Mach 1. Battling with supersonic airflow below mach 1 is difficult and de-stabilising. We will be experimenting with dropping light airframes with a camera at the nose. Before we reach controlled airspace, we will deploy our parachute and have a reserve one for safety. We will watch carefully to see the effects on stability.

The diagram below is one solution to moving the mass required for stable flight in both modes. The pump must be fast and the liquid must stay “thin” and not become viscus. We will need baffles to slow the sloshing around during the changeover. These divide the tanks into chambers with some small holes joining the chambers to allow them to fill.

Centre of gravity adjustment transitioning from dive to level flight

There are other solutions such as screw thread that will shift the battery and electronics forwards or backwards. Since the flight is short the transition only needs to be one way, the design is thus simplified. I am not a fan of shifting the battery and electronics around. It will take a large movement to have the desired effect and it could cause wired t break if they get caught on something. I personally favour pumping the fluid from forward to back as shown above. Moving it down during horizontal flight creates even more stability by creating dihedral effect between the wings on an otherwise symmetrical aircraft.

Dihedral in aircraft is the inclination of an aircraft’s wing from the horizontal, especially upwards away from the fuselage. in this case it is the centre of gravity that I am measuring it against and this indicates that the weight is below the wings and the aircraft will be easier to fly.

Below is another thought on using systems, but this time we vent the fluid without the need for a pump.

gravity does the work for us and we remove the liquid away from any potential problems within the aircraft. Making it lighter will also make it more controllable once out of the dive.

gravity does the work for us and we remove the liquid away from any potential problems within the aircraft. Making it lighter will also make it more controllable once out of the dive.

Whatever system we chose, we will be writing it up here. we need to fly the craft and we also have access to a wind tunnel for subsonic tests.