The earth is a whirling disc in front of the nose. And I am due to die in perhaps 30 seconds because I seem unable to stop it rotating in this manner.

Christ, the brief was simple enough. Go fly a Certificate of Airworthiness test flight on this very ordinary single-engine Cessna. Here’s the clipboard, just fill in the boxes…

The clipboard is now… somewhere, and I do not care where. The only thing I care about is the whirling ground filling the windscreen. A few seconds ago I kicked the Cessna into a ponderous spin to the left, let it rotate four turns as per the flight-test form, and then recovered.

Except that this spin turned out to be not that ponderous. And it hasn’t recovered.

The Cessna and I spin on. Suddenly I have become religious. Jesu, I have full right – out-spin – rudder on. I’m pushing forward on the yoke… but the spin goes on.

In desperation I try an old and extremely dubious solution – rocking the stick and throttle back and forth together. Normally an old wives’ tale…

But this time it works.

It works not because of any aerodynamic magic, but simply that I suddenly realise the throttle lever is physically moving more than the control yoke. I shove the yoke forwards two-handedly with all my strength…

And the spin stops within a half-turn.

That was 40 years ago. But the moment has kinda lived with me.

Why am I telling you this? Well, bear with me. BLOODHOUND is unashamedly – indeed most proudly – an engineering adventure. All engineering adventures hit snags from time to time. Mostly they keep quiet about them – but BLOODHOUND does not. The BLOODHOUND adventure is out there for all to see – snags, warts and all.

And recently BLOODHOUND hit a new snag. Very much related to that spinning Cessna and a very frightened Brian four decades ago.

This is in fact a tale of two centres. The first centre is called Centre of Gravity – C of G, or CG. (Modern scientists tend to call it Centre of Mass, which I’d always rather regarded as being the property of the Catholic Church – but there you go). The second centre is Centre of Pressure – C of P, or CP. The balance between the two in a moving object is a bit like the scales of justice – it has to be right.

Which it ain’t quite today in BLOODHOUND. Just as it wasn’t quite for me on that interesting day 40 years ago in the Cessna.

Let me try to explain. Please be tolerant of this endeavour, because I is a simple pilot, not an aerodynamicist and most certainly not a mathematician. On the credit side this means I am not going to throw complex formulae at you, but on the not-quite-so-credit side, if you yourself are a highly qualified design engineer you’ll probably be better off turning to the sports results or the cartoon page, because I am going to sound a bit kind of kindergarten. In fact whenever I talk about technical issues in front of the BLOODHOUND team I can never quite rid myself of the suspicion that if I turn round quickly enough I’ll catch a row of the sort of indulgent smiles parents wear when they’ve shoved little Johnnie out on the stage to recite The Boy Stood On The Burning Deck.

Anyway. We’ll start with Centre of Gravity – CG. Imagine you have a playground see-saw 3 metres long, pivoting in the middle. You have a beefy kid full of roadkill-burgers on one end, weighing 50 kgs. (The kid, not the burgers). On the other end you have a skinny kid weighing 25 kgs. Clearly the see-saw will come to rest with beefy kid on the ground and skinny kid on the upside, unless beefy kid has already shot him into the air like a round from a siege catapult. Okay. To make it balance you could either put a second 25 kg kid on the skinny kid’s end – or you could move the see-saw board along the pivot so that beefy kid was one metre from the fulcrum while skinny kid was two metres. Skinny kid would then have twice the moment arm of beefy kid, and so they’d balance. The new pivot-point would be the Centre of Gravity – the point where all the weight comes together in balance. If you bolted a hook onto the fulcrum point and lifted the whole thing into the air you might get a pair of brown-trousered kids but there would be no change in balance – that’s the definition of CG.

Now CG does not change with movement – acceleration, roll, pitch, yaw… nothing. I admit that I’ve always had trouble getting my head around this, but then I also admit to having trouble getting the top off some fashionable shampoo bottles, not that I try many fashionable shampoo bottles. But CG does move with internal weight-shift. If all the passengers on an airliner suddenly decide to rush down the back then the pilot’s going to know about it quick-time. And as fuel gets used up so it usually shifts the CG at least a bit and he knows about it slow-time. Thoroughly expected.

 

Losing a fifth of your weight in 40 seconds…

Such CG shift being a fairly big factor in BLOODHOUND. There won’t be any passengers rushing down the back, since (a) there ain’t no room for stowaways, and (b) in any case given the heat at the back end, incinerated stowaways probably don’t weigh all that much anyway. But BLOODHOUND is going to lose about a sixth of its own weight in fuel-burn in 40 seconds during acceleration – a huge percentage of mass change in a very short time. It may very well be highly environmentally-friendly fuel – but the car’s a lot more interested in what it does to the CG. And what it does is moves it backwards…

Okay. Now we have to leave CG for a moment, and come to CP – Centre of Pressure.

Any moving vehicle obviously has airflow over and around it. Some of it at higher pressure than ambient – such as a car ramming its snout into the air – and some of it at lower pressure, where the airflow has been speeded up.

So, in the same way that CG is the centre-point of weight, CP is the centre-point of all aerodynamic effects. So far so good.

Now let us relate this to stability. In any moving vehicle the centre of weight, the CG, wants to keep going straight on at all times. The CP, on the other hand, is not in the least bit picky, being merely a resultant of pressures. If the CG is ahead of the CP then the CG effectively ‘tows’ the Centre of Pressure, so that what the scientists call any ‘perturbational’ effects – changes of direction in any plane to mere mortals – tend to be damped out unless maintained by human or other agencies who wish things to remain perturbed. Said vehicle is ‘statically stable’, or ‘positively stable’. A bit like throwing a dart forwards, as BLOODHOUND’s Brian Coombs, that most eminently practical engineer, puts it.

If the CP is in front of the CG, however, the CG tries to overtake the CP. So any ‘perturbation’ gets magnified instead of damped. Sometimes diverging very rapidly – a bit like throwing the dart backwards and seeing how far that gets you. This is ‘static instability’ and is generally speaking Not Good For You.

In between the two is ‘neutral stability’, where the CP is pretty much in the same place as the CG. Then the thing doesn’t tend to self-correct but also doesn’t tend to diverge either, just staying in the new attitude you’ve put it in. This might sound useful in certain applications, but is in fact highly vulnerable to the factor which towers over all CG / CP calculations and predictions of any kind.

CG is static apart from load alterations.

Whilst CP is most definitely NOT static.

CP moves around with changes of airspeed, changes of angle of attack of a wing or anything else, pushing out airbrakes – you name it, the CP moves. Increase the angle of attack – the CP moves forwards. Increase the speed – the CP moves backwards. 

 

Back to the Cessna

Which brings us back to that Cessna 40 years ago. Because this was an air-test, I’d carefully loaded up the thing with full fuel and a couple of sandbags in the back to bring the weight up to max gross. I’d equally carefully gone through the weight-and-balance graphs to make sure I was within CG limits – which I was, although right on the aft limit. But the aft limit’s still within limits.

CG and CP however moving closer together…

I was used to spinning the Cessna with a student in the other front seat – a much more forward CG. In that configuration the problem wasn’t recovering the spin, it was holding the aeronautical clown in the spin long enough to demonstrate recovery, since if either of you burped the blasted thing fell out of the spin of its own accord anyway.

With aft CG on that day it simply had never occurred to me that I’d need more pressure to move the elevators to shift the CP backwards. A lot more pressure…

Simple as that.

Now in aeroplanes or cars the further forward the CG is from the CP then the more stable the vehicle. This is usually more or less desirable but not always – F1 cars, jet fighters and very high performance aerobatic aircraft being prime examples. Here you want CG and CP to be close together and only just stable, to decrease control pressures and increase manoeuvrability. And in top-class aerobatic aeroplanes especially, the CP moves around like a bar of soap in a Jacuzzi, frequently darting well off to one side or the other, never mind fore and aft. In several aircraft I’ve had, the controls – particularly the major stabilizing ones, rudder and elevator – would ‘lock’ on full deflection during certain highly gyrational manoeuvres. Not lock as in being immovable, but as in being perfectly happy to remain at full deflection if you took your hands and feet off, leaving the aeroplane in an interesting rotation which it would not of itself recover from.

 

But there is a but…

Sounds worse than it is, because with a bit of muscular heave you could overcome it, effectively moving the CP backwards and re-stabilising the thing. In an aerobatic machine this transition from stable to unstable is thereby acceptable, within reason. You can recover it, yes. But there is a but…

And that but is that actually controlling it in the unstable regime is berluddy difficult, because everything you do tends to perpetuate itself. You are not pushing a control and holding it – you are nudging it and then trying to hold it back, with no natural ‘feel’ whatsoever to give you a clue. You are trying to fly the dart backwards. Not totally impossible – but very, very difficult.

A small handful of modern jet fighters are deliberately unstable at all times, and thus require a computer to interpret the pilot’s intentions and move the controls. If the computers go phut the standby system is extremely simple – the pilot ejects, right now and no bloody argument.

So the CG / CP relationship of any vehicle is kind of important. And especially if that vehicle happens to be a car which is intending to punch through the transonic speed range and deeply into the supersonic.

Such as BLOODHOUND.  In fact, uniquely such as BLOODHOUND.

Firstly BLOODHOUND has to be stable in pitch, so that over the huge speed-range it never becomes either a space-shot or the fastest tunnelling machine in history. This is the really difficult bit aerodynamically, and the answer has taken nearly three years of research because this was the first ‘must-have’.

However…

To say there is one Centre of Pressure at any given speed is perfectly correct. (Kindergarten lecturer takes sneaky glance backwards to look for any parental pursed lips). But the fact is that the one CP position is a total resultant of at least three parameters – pitch, yaw and roll.

And it’s perfectly possible that while the total CP might look fine, any one element can be out of limits in its own plane.

And BLOODHOUND’s calculations, after the Holy Grail of defeating rear-end lift was finally solved, inexorably transpired to put the yaw CP ahead of the CG.  To use the scientific term, the Yaw Static Margin – basically the distance between the CG and the CP – turned out to be… er, well, a negative value. The wrong way round. The CP ahead of the CG.

Which makes the car naturally unstable directionally. Like trying to fly the dart backwards. In a way, if you like, like trying to break the Land Speed Record driving an articulated lorry backwards. Now Andy Green’s certainly good, but nobody ain’t that good…

 

You discover a problem…

Well, that’s an engineering adventure for you. You discover a problem… you solve it… then you find another problem and solve that…and then you find that the solutions have aggravated yet another problem…

In a way there was a certain inevitability about it. The major weights of BLOODHOUND – the jet, the rocket, and structure associated therewith – were always going to be down the back, obviously. The problem was that until you could get on with highly detailed internal design you didn’t know just exactly where down the back. And you couldn’t finalise highly detailed internal design until you knew the final shape of the car. And you couldn’t know the final shape of the car until the fundamental aerodynamic issues were resolved…

So the not-unknown BLOODHOUND circular parade, then – aerodynamics lead the band, and everything else has to dance to their tune.

Then, when you have got everything finally located, working out that very final CG. Which is not exactly a 10 minute job.

Weight and balance calculations are part of any pilot’s stock-in-trade, of course – but as an aviator all you’re really doing is plotting-in the variables, usually little more than weight and location of fuel, crew and payload, to ensure that (a) you’re not going to be over gross weight, (b) you haven’t put that consignment of gold bars ‘way down the back so as to drag the CG backwards beyond its aft limit, and (c) that the CG’s going to stay within limits even after you’ve burned up fuel.

Mention this to design engineers – especially BLOODHOUND design engineers just at the moment – and they will at best give you the sort of wearily tolerant half-grin normally reserved for the family puppy who has just made a large puddle on the carpet. Because they have to do CG calculations not just on the variables, but the whole cotton-pickin’ assembly. Every detail. Oddly enough really big ‘imported’ items like the EJ200 jet are not the biggest problems because they come with their own weight and balance figures. But everything – everything – else, you have to calculate.

Say you have a thrumblecrudgit two metres long as part of the brake chute mounting. You have to calculate its weight, then calculate where its own CG is as an individual part, then insert the resultant into the overall CG calculations, which are ongoing and by now taking on the size of a substantial novel. Then someone comes along saying we need a bigger bolt at the back end of the thrumblecrudgit, so that alters the thrumblecrudgit’s CG slightly so you have to re-calculate that and then re-insert it into the overall equation. It would of course be easier to kill the harbinger of bad news, but since there are hundreds of these happenstances in any given period the pile of bodies outside the BLOODHOUND Technical Centre would be bound to be noticed, and would anyway start to niff, what with our lovely warm spring.

Which is why the final CG figures are not exactly easy. You have regular reviews and then a final review and then a final final review – which lasts just fine until something else changes and you need a final final final review. Which is why the definitive results tend to come late in the day. And why, as any aircraft or car designer will tell you, one of the most attention-grabbing moments in the whole process is when the prototype is rolled out and finally weighed end-for-end, thus proving or disproving the calculations. Oh, you can stay more abreast of the shifting CG sands if you’re an F1 racing car designer, because for one thing you’re working from a known starting point, and for another you won’t be one F1 racing car designer anyway, but one of a team of maybe 180 full-time designers – which believe it or not is about average for an F1 team. Year round. This provides a certain attention-to-spare capacity which the BLOODHOUND design team, being less than a tenth of the size, does not actually quite have to hand.

The final Centre of Pressure is equally elusive for much the same reasons – you have to wait on the primary aerodynamic outcome.

Anyway… the net result of reviews in the past few months indicated that the caboodle was the wrong way round. CP ahead of CG. Quoting detailed figures for this is fairly pointless because they change pretty much daily as the team find the solutions. But roughly speaking they are aiming for a 3% Yaw Static Margin. This decodes as the CG being 3% of BLOODHOUND’s wheelbase ahead of the CP – which amounts to some 27 cm. That’s the objective.

Which meant either shifting the CG something like 75 cm forwards, or the CP something like 75 cm back. Or some sort of combination of the two.

On the plus side the CP moves backwards with speed-rise, reaching its aft-most point at about Mach 1.1.

On the minus side the CG also moves backwards with speed-rise, mainly courtesy of a tonne of HTP going missing in 20 seconds.

So what are the answers? Much work has already been done, and there is definitely light at the end of the tunnel which will not prove to be an oncoming train. The solutions are coming in steadily.

(In fact even as I write this I learn that the latest iteration – yet to be fully proven, of course – has moved the CP back to about the same position as the CG. Not the final answer of course, but moving, moving…)

But t’ain’t simple. Several obvious solutions turn out to be not entirely virtuous circles. Yes, you can move the yaw CP backwards by moving the fin rearwards and making it bigger – which it now is, by some 70% over the original area. You can also re-visit the shape of the fin, because different shapes react differently to Mach number rise.

The non-virtuous of that is (a) you’re also moving the roll CP upwards – acceptable within limits, but not top of the Christmas card list – and (b) you need a heftier fin structure. Which inevitably weighs more. And that accordingly shifts the CG backwards at the same time…

You can add a strake to the main fin. This will move the CP a tad further aft in its own right, will add strength in its own right, and also enable you to stream the airflow at the base of the fin so as to allow a thicker fin-root aerodynamic section without increasing drag. Significant advantage, because the broader the base of the fin is, the easier – and lighter – the structure needed to transfer the loads.

And you can add a central underside strake and a couple of small ventral stabilisers – that’ll shift the yaw CP backwards a bit more.

And you can also alter the shape of the front end slightly. As a change this may sound as if its middle name is ‘drastic’, but in fact not necessarily so. When you go rearwards from the pointy bit at the front of a supersonic vehicle (or even a high-speed subsonic vehicle, come to that) you eventually reach a point where the cross-sectional area is at its fattest. Plucking a figure out of the air, let’s call it 4 square metres. Thereafter as you move further back down the vehicle you want any cross-section slice to also be 4 square metres, at least until you reach another pointy bit at the back, if applicable. This is known to designers as Area Rule. You can’t always achieve it because other designers will insist on adding fripperies like engines and wings – but you don’t want the area of the cross-sections changing if you can possibly avoid it. But area isn’t the same as shape. You can change the shape, within reason, without changing the area…

Naturally, it is not as simple as this, and naturally I understand it the way I understand a Beethoven Symphony, which is sadly not a lot. But what it comes down to is that BLOODHOUND’s fuselage is currently pretty much slab-sided from a bit forward of the front wheels right to the back. So any cross-airflow aft of the CG will help to stabilise the car – but any cross-airflow forward of the CG will fight against that.

Answer – reduce the slab-side forward of the CG. Kinda round off the sides a bit so that a cross-airflow has less affect at the front than it does at the back, while keeping the same area. A slightly puff-cheeked BLOODHOUND.

Not by any means impossible to do. And being looked at along with the other solutions.

 

And the other way round…

You can also look at it the other way round and shift the CG forwards. You can put the ice-cooling system for the Cosworth engine as far forward as you can get it…

Already done that.

Or you could put the whole Cosworth-HTP pump assembly forward of the cockpit…?

Drastically complicated, but never mind that. Forget it right from taw ‘cos there ain’t enough room for it anyway. This is the pointy-end, and so is kinda cramped…

You could put the computers up the front end – fine, but being all high state-of-the-art they’re featherweight anyway. You can put things like Nitrogen pressure bottles and hydraulic accumulators up the front – but they also are not exactly heavy.

Or, as I suggested with one of my brilliant flashes of insight, you could always stuff a couple of bags of concrete right in the front of the nose-cone…

D’you know, it’s amazing the burning feeling you get when ten pairs of eyes on the BLOODHOUND team focus on you like lasers…

The fact is that the snag is cornered. The solutions are being honed and polished, but they’re not far away. And I don’t think they’re gonna involve a couple of bags of concrete…