Unlikely as it may sound, an engine test cell frequently has something in common with a film unit at work. A similar apparent mess of cables behind the scenes: a similar coming and going of expert heads shaking or nodding wisely at the latest proposed solution to the latest discovered problem.

And above all, the hanging about. Oh Gawd, the hanging about. On more than one film set I have been left sitting in an idling aeroplane while a Director bawls something like: “Get a scrim on that brute – now!”  I have not the faintest idea what a scrim is – possibly something an all-female rugby team does – but have a vague notion that ‘brute’ might refer to a huge great spotlight which is pointing at the cockpit just in case the strong midday sun should suddenly decide to nip off to McDonald’s. The one thing I do know from experience is that the ‘scrim’ (whatever it is) will inevitably be at the bottom of some lorry’s equipment pile, and that ‘Now’ means at least a half-hour wait. I shut down the engine. As expected, this provokes the Mighty Director to despatch a minion to the cockpit and shout at me: “Why have you stopped the engine?”

“Because it’s going to melt if it gets any hotter on the ground”.

“Melt?” He looks at the engine cowling as if expecting molten aluminium to be pouring out of the bottom. “It’s not melting! Mr (XXXX, a famous film director) wants the engine running!”

I can be tactful at times. And this is one of those times. I smile sweetly and say gently: “I am not going to over-temp this engine. Please tell Mr XXXX to go take a long walk off a short pier”.

For some reason this does not go down too well. The Mighty Director himself storms over to the cockpit and delivers himself of a truly impressive salvo of swear-words, many of which rhyme with ‘duck’ but some of which are quite novel even to my ears. I listen to the tirade, trying to make mental notes of the new words, then explain soothingly that (a) I am not prepared to overheat a perfectly sound engine, (b) that nothing is happening anyway, and (c) that both he and I know perfectly well that nothing will happen until the lads have found the scrim (whatever that is) and put it on the brute (if it’s what I think it is). Which will take at least half an hour. And then I’ll start up again.

To my surprise the Mighty Director suddenly grins and says; “Oh, you’ve learned about filming, then!”  And stalks off to shout at somebody else.

Well, here and now in this engine test cell at Cosworth’s Northampton factory there is no shouting or swearing whatsoever.

But there is still the hanging about.

Now please don’t get me wrong. I have been in many engine test cells over the years and the hanging about is endemic. It comes with the territory. And especially so if any part of the engine or systems is new and untried. The whole point of testing an engine or system is to discover snags – indeed, even to deliberately create snags at times – and then fix ‘em. Mostly this means maybe two people wielding spanners at the problem while the rest of the crew… well, hang about. Nothing else they can do. Especially if they are supernumeraries like me, who are not going to be allowed to touch a spanner and in my case have also been adjured in no uncertain terms not to even think hard at anything which even faintly resembles an electronic component. Which seems to be ‘most everything. Clearly, word of my computer non-skills has gone before me.

Which at least gives me time to study this most extraordinary test-bed. In the whole of Cosworth’s spread-out industrial complex there is no other test-bed remotely like it. In fact, come to think of it, there is no other test-bed remotely like it on the entire planet, unless somebody is keeping awfully quiet about something.

At the front of it is an unprepossessing electronics box, containing, among other things, a small black box and a small silver box. Dr John Davis, BLOODHOUND’s guru of electronic and hydraulic control systems, explains, bearing in mind the mental stature of his one-man audience and therefore avoiding long words.

“The black box is the engine’s Electronic Control Unit. The silver box is part of BLOODHOUND’s control system, and talks to the black box to produce the engine control. Okay?”

Like any high-time pilot I have an intelligent answer ready for this kind of situation.

“Aahh”, I say.

Behind the box is the Cosworth CA2010 engine, capped with a large red notice saying ‘AVS OFF – DO NOT TURN OVER’. AVS stands for Air Valve System, because the inlet and exhaust valves are closed not by springs but by air pressure at about 20 bar – somewhere around 300 psi in my old-speak. (See Tobacco, Flesh and Testing, From Outside the Box #26). This pressure is provided by an incredibly small compressed-air bottle – but if you turn the engine however slowly with AVS off you could quite easily do £250,000 worth of damage because the valves won’t be closing. I resolve not to turn the engine under any circumstances. Not that I would know how to anyway, because no normal controls are in evidence. We’ll come back to that.

 

Forget the radiator

Above the front of the engine is a cylindrical tank. This is the ice tank for engine cooling – because BLOODHOUND will not have a radiator, on account of radiators needing air-intakes, and air-intakes mean aerodynamic drag. Hence the ice-tank, which is far, far smaller than I had imagined, being maybe 80 cm long by about 30 cm in diameter. I look around for a bucket of ice cubes.

“Where’s the ice?” I ask John.

“We don’t need it today. These are short runs”.

“Oh”. So can this engine run for maybe an hour or so with no radiator and not even ice to cool it?

I have a lot to learn. I get a clue with my next stupid question. Pointing at a small vertical cylinder just aft of the (today, non-ice) tank, I ask John if that’s a coolant expansion tank.

“No. That’s the fuel tank”.

Hmmm. Well, moving on…

At the back end of the engine a short shaft leads to what I can only describe as a transfer box. ‘Tisn’t a gearbox – the team’s still waiting for that to be completed – so this is a fabricated… well, transfer box. For test purposes only. It houses one of those amazingly tiny race car clutches – a multi-plate affair with a diameter of maybe 13 cm (5 ins) – plus a simple dog-clutch to engage Daniel Jubb’s incredible Stentor-based HTP pump located right behind the transmission. (See From Cold War to Car, From Outside the Box #23).

There is however a drawback to this transmission box – namely, that it perforce can only be one-to-one, so the pump will be running at the same rpm as the engine. Whereas what really needs to happen is the pump peaking at 12,000 rpm while the engine peaks at 800 bhp at some 18,000 rpm – a gearing ratio of 3:2. But the gearbox isn’t around yet…

“I’d like to get the engine up to 12,000 rpm, but we’re not going to manage that on a gearing of one-to-one”, says John matter-of-factly.

Intellectually I know the figures perfectly well – but seeing the set-up in the flesh sort of brings it home to a chap. All of that Cosworth engine to drive one innocuous-looking pump with an 18 cm (7 ins) diameter rotor inside it – and it won’t do it properly until the gearing is right. From the point of view of this early test programme it doesn’t matter too much, there being plenty of other parameters to be getting on with. But all the same…

“The engine clutch and the pump dog-clutch are quite complex”, John is saying. “During engine start-up I have to de-clutch the engine clutch, engage the dog-clutch, and then re-engage the engine clutch”.

I blink. “You’re gonna be busy”.

“Oh no. I’ve got four seconds for that”.

I shut my mouth. Clearly something going on here I don’t know about. Not that that’s any kind of first.

Also on the transfer box there is something else I hadn’t expected to see – a neat little lightweight starter-motor, rather like the ones we had on aerobatic aircraft. But Formula 1 cars don’t have integral starter motors – they’re spun up with an external starter. I open my mouth to ask – then shut it again when I realise the obvious answer: viz, how could you possibly engage an external starter to an engine tucked away in the entrails of BLOODHOUND?

Behind the pump, and conjuring up a most unlikely vision of an adolescent elephant having strolled into a Cosworth test cell and there leaned peacefully on the nearest convenient structure, is the massive HTP tank. (See - The Importance of Slosh and Slam, From Outside the Box #29) This is propped up at an angle of 63 deg, to simulate the G force of acceleration.

The ultimate idea is that the whole kit and caboodle – engine, transmission, pump and tank – will be similarly hoisted up to 63 deg to make sure no component starts sulking under full power and full G.

However there are two problems to this.

One is that the engine in this test cell is an ‘ordinary’ Cosworth – insofar as any F1 engine has ever been ‘ordinary’. The ‘real’ BLOODHOUND engine is having its oil system modified to cope with the G, and will be with us shortly. So this particular engine remains horizontal.

 

Inventing the staircase

The second problem is a wonderful example of how plain logistics can thwart the most determined intentions. In Dr Who’s long confrontations with the Daleks all he ever really had to do was invent the staircase, and they’d have been intoning “Exterminate…Extermi -  oh sh** I’m falling down the steps…!”  Thwarted. The equivalent here is that if you tilt the whole caboodle up to 63 deg – then you’re simply going to hit the roof of the test cell. Oh sh… I mean what a dashed nuisance. Somebody will sort it out, of course – but at the moment people are looking at the problem thoughtfully.

I go back to the tank. On top of it is a Christmas tree of pressure control do-daahs, pressure-relief thingummys, and a large ball-valve unit which will act as the filler-point when BLOODHOUND is replenished from an HTP bowser. Plus sensors to tell you a great many things, including for all I know when your egg is soft-boiled. The tank is also surrounded by four nitrogen bottles roughly the size of oxy-acetylene welding bottles – which means as tall as me. And each one probably heavier than me. These are required to pressurise the HTP tank to 24 psi and keep it so pressurised throughout the run despite the fact that the gas-space in the tank will be increasing something like ten-fold inside 20 seconds.

“Er, John. Biggish sort of tanks, these. And heavy. How can you get anything like this into BLOODHOUND?”

“Into BLOOHDOUND? Of course not. The nitrogen tank in the car will be maybe 30 cms (12 ins) long and pressurised to 800 bar. That should last for two runs. These are just for these tests”.

Aahh. Of course…

 

A little trailer…?

I try to imagine driving a vehicle – any vehicle – containing a pressure vessel – any pressure vessel – pumped up to 800 bar, which is getting on for 12,000 psi. I shall have to get used to the idea of course if I’m going to continue with my plans for kidnapping Andy Green and driving BLOODHOUND myself – but 12,000 psi, forsooth!  I find myself kinda hoping that particular vessel will be as far away from the cockpit as possible. In fact preferably towed on a little trailer behind the car, although I do recognise this may not be entirely practical.

Well, okay. Continuing on round the test bed I step over electrical-hydraulic umbilicals leading to an external hydraulic pump driven by an industrial electric motor. I raise my eyebrows to John.

“Just for these tests. F1 cars have a small but very effective engine-driven hydraulic pump, which runs just the steering, gearchange and throttle. We need more pressure so we’ll have two of them, one on the engine and the second running off the gearbox”.

“Aahh…”

Carrying on, I duck under the HTP pump outlet pipe. In the car this will go straight to the Falcon rocket. On the test bed it is a most substantial flexible hose of about 7 cm (2.5 ins) inside diameter rigged upwards and splitting into two in a Y junction, the outgoing hoses both emptying into a couple of huge translucent polythene tanks mounted high up in the test cell on a steeple of pallets.

High up? Why high up?

Well, partly to compensate for the fact that the test rig here is pumping DI water (de-ionised water) instead of the HTP (High Test Peroxide) which will be used in the real thing. HTP is considerably more dense than DI water, so pumping DI water up to a height kind of reproduces the pressures. A bit. There is no chance of using actual HTP in these trials – HTP tests are for a firing range, not for here at Cosworth.

The other reason for the high-level tanks is much more mundane. DI water is not hugely expensive – but you don’t want to throw it away. And the HTP bowser with its integral high-speed pumping system does not yet exist. So in the here and now, once you’ve emptied (or nearly emptied) the HTP tank you open the taps of the polythene tanks and let them drain down into two large plastic drums on the floor. Then you pressurise the drums with an air-line so that the water flows out through a couple of pipes which feed into the top of the HTP tank.

This is, needless to say, an improvised system. Well, all right, a bit of a lash-up. It doesn’t impinge on any of the kit under test, so it doesn’t actually matter – but it does make replenishing the HTP tank a lengthy and sometimes slightly damp business. In fact two hours lengthy if you’re starting from scratch, and a process which involves leaping up ladders to the top of the tank to check on the current state of the nation.

However, this is not the only thing that can take two hours. I’ve mentioned before that F1 engines are kinda picky in the matter of temperatures – and round the far side of the engine is the F1 cross between a witch’s cauldron and an electric kettle. It consists of a round tank with a heating element, and an electric pump which circulates hot coolant around the Cosworth’s galleries. This gradually brings the temperature up to 70 deg C, and by heat-soak the oil up to 45 deg. All F1 engines use much the same system of external pre-heating, which of course is un-plugged before engine-start. To me there is a certain irony in expending much technology heating it up, and thereafter, once it’s running, much more technology cooling it down again.

I do not mention this. All the people around me – very much including John Davis – are F1 technicians to their bones, to whom this procedure is absolutely normal. Any smart cracks from me will be received by a row of patiently blank faces.

“We’re ready to run”, says John. “Come into the control room”.

“I’d rather stay here. I’d like to hear it first-hand”.

“Forget it. No-one stays in a test-cell while an engine’s running. Come to the control room”.

I go to the control room. As I enter, blast-doors thump shut behind me.

I look around in awe at a sci-fi room out of Star Trek. There are maybe a dozen live screens on one artificial wall. And another five or so on a bench in front of a long window looking out at the test-bed. On not one of these screens is a presentation I remotely understand.

John flicks a keyboard and the display changes on one of the two screens in front of him. “This…” he waves a hand at it, “…is my control sequence. This is what talks to the engine ECU”.

 

You want to run a space station…?

On the screen are a dozen or so shapes faintly resembling thinks-bubbles in a strip cartoon. Each bubble contains several lines of tiny writing or hieroglyphics, and there are connecting lines running every which way between the bubbles. For a moment I think I’m seeing a much-enlarged picture of amoebas in mid-orgy – but after blinking at it for a minute I realise this is some sort of flow-chart of actions. A flow-chart of such complexity that you could probably run a fairly substantial space station with it.

John clicks a mouse and the screen changes to menu boxes. Click again and sub-menu boxes appear. John selects several in turn and types rapidly.

I make one of those intellectual leaps for which I am most justifiably not famous.

“So you’re programming the whole run?”

“Yes”.

“So for that clutch-out, clutch-in, clutch-in sequence in four seconds for the pump, for example – you don’t have to press buttons to make that happen?”

“No, no. I programme it in”.

“And you designed this whole programme?”

“Yes”.

I mentally say “Aahh” again. John is busy on the keyboard, and I’m now getting a faint whiff of the Herculean task he has achieved in the few months he’s had working alongside Cosworth.

John leans across me to a young Cosworth engineer called Jules, who is manning his own screen plus holding a remote control which includes a red button.

“This is a run to check remote-start remote-kill”. John’s voice reminds me of my own when I used to call: “Display briefing!” and everybody snapped-to. Jules nods.

“Any reason why we should not run?” John looks around the room.

Shaking heads. Jules says: “No”.

“Starting”.

John presses a single key on his keyboard and the engine starts. I’d like to tell you there’s a chirr-chirr of a starter followed by a yelp as the engine fires up – but inside the soundproof control room all I can hear is a whirr like a distant electric motor. It runs… the DI pipes jerk and slosh water into the tanks… for a second the revs rise audibly…

 

What’s bust?

And then everything stops. Just like that. As if an axe has come down. Clearly the engine has blown up or something else has broken…

“Run time?” says John coolly.

“Eight seconds”, says Jules.

“What’s bust?”, says I.

“Bust?” says John. “What do you mean, bust?”

“Well, er… the engine quit, didn’t it? Something wrong?”

“No, nothing wrong. This was just to test remote-start remote-kill. Jules here has an independent back-up kill button in case my system failed. But it worked fine”.

“But… the engine stopped dead…”

“Certainly. An F1 engine will stop in 0.4 seconds from 18,000 rpm. How else can you get the really fast gear-changes?”

Aahh…

Less than half a second from 18,000 to stopped. I close my mouth and try to wipe the bemused look off the front of my head. John and Jules study their computer screens, which now show graph-type presentations with perhaps a dozen lines of different parameters plotted on them. I long for an old-fashioned round rev-counter and round temp and pressure gauges…

John has found a problem among the graphs. The HTP pump outlet ball-valve wasn’t opening fully. In a characteristic gesture he tosses off the glasses he uses for computer screens, slaps the bench in a quick drum-roll, and strides into the test-cell.

He and a Cosworth guy called Chris remove the ball-valve, strip it as far as possible, can’t find anything wrong, and put it all back together. John ups the hydraulic pressure to open the valve.

Back in the control room John briefs again. “This run is to 7,600 rpm. Any reason why we should not run?”

Heads shake. John taps the keyboard again. The engine – heard from inside this soundproof room – whirrs into life, stabilises at 4,500 rpm idle for a few seconds, then shrieks almost silently up to 7,600 for a moment. The delivery pipes pulsate, DI water sloshes into the tanks at an amazing rate…

And then everything stops again. Instantly. But this was a long run. 12 seconds, to be exact…

That delivery ball-valve has still not behaved. John says the pump pressure has overcome the hydraulic pressure. He does his drum-roll on the bench again, then paces out into the test-cell. Wherein he locks the ball-valve to fully open, shelving the actuation issue into a file marked: ‘Must solve – but can live with for now’.

After the third run most of the DI water has been sloshed into the overhead tanks, so there is an hour’s hiatus while it’s drained down and pumped back up into the HTP tank.

On the sixth run the rpm peaks at 7,500, producing an astounding cataract of DI water. About one quarter, as I understand it, of the pump’s capacity. I am beginning to get the hang of this. The engine fires up, idles for a few seconds, picks up to maybe 6,000 rpm to stabilise temperatures for another few beats of time, then boosts up to target rpm for maybe a whole two seconds – and then stops. That last second or two is enough to harvest the data…

“7,500 rpm used 54% of throttle”, says Jules. “Any more and the engine’ll bog down – too much load at low rpm”.

John goes through his glasses and drums routine again and exits to the test-bed to make a change to the pump outlet system.

By the end of this day the engine has run seven times – for a total run-time of less than three minutes. But lessons have been learned and solutions are suggesting themselves…

That is life in a test-cell.

It is called real-world. Where everything is not necessarily perfect, and does not necessarily work first time.

 

Postscript

All of the above is just a snapshot over one day. Over the next two days more runs follow. John changes the pump outlet pipes to smooth bore, which helps quite a lot. He also re-programs the clutch so that instead of engaging at idle he can raise the engine speed to 10,000 rpm and then slip the clutch in gently – ‘gently’ being over maybe two or three seconds. Slipping a clutch at 10,000 rpm sounds barbaric to me, but apparently in F1 life it is not unusual. John calls it ‘launch control’. And this enables him to get the pump-speed up enough to produce 420 psi of pump delivery pressure – a bit over a third of what is required. In the last 1% of clutch engagement the revs drop to 7,000 as the full load comes on.

This is as far as John can go with the direct-drive rig. The gearbox won’t arrive until October, but in the meantime he is creating ‘Plan C’ – a new transfer box with the necessary 3:2 down-gearing, which will be ready in a few days. John stresses that this is an interim measure and ‘May not have a long life’.

This is real life in a test-cell.