The Rolls-Royce and Electroflight Spirit of Innovation smashed the world speed record for electric aircraft. Mark Greenfield worked with the team throughout and brings exclusive insight
Words: Mark Greenfield Photography: Rolls Royce and Electroflight
18 January 2022
In photos the aeroplane looks dramatic – futuristic, sleek, somehow disproportionate. In the metal the impact is more extreme, more space age. The engineering quality is outstanding and the whole aircraft exudes an almost menacing sense of brute purpose. While functioning as a technology breakthrough (FLYER May 2021), this is very clearly a machine that is designed to go fast. Very fast indeed.
The record for all electric aircraft flight had been set by a Siemens eAircraft-powered Extra 330 LE in 2017 at 213mph. The Spirit of Innovation increased the record on 21 November 2021 to a massive 345.4mph, at the same time becoming the fastest all electric vehicle on the planet. Rolls-Royce is hoping the record will be ratified imminently.
I had worked with the Rolls-Royce pilot team on Upset Prevention and Recovery Training, and three years ago the Flight Upset and handling challenges of a little-known prototype aeroplane were inevitably a strong focus of attention. It was agreed that converting the pilots onto our Extra 300 might be a useful workup in flying a powerful and lightweight taildragger with poor ground visibility. In anticipation of flying an aircraft with eight minute endurance at full power, it was also of value in practicing forced landings at the high glide speed (130kt) required by a wing designed with the sole intention of going fast. The Extra was also used as a safety chase aircraft on the early flights, a decision that initially compounded operational challenges, but later proved to be very much worth the complications.
The first flights of the new aeroplane were initially anticipated at the beginning of 2021 but, unsurprisingly for a project of this nature, the timeline shifted to the right.
So in the middle of July 2021 I found myself landing the Extra at Boscombe Down, reminding myself not to flare high because of the wide runway (and nearly succeeding), to work on the practicalities of using it as a chase plane. Initially, a P-51 Mustang was used as a stunt double and the integration of record aircraft, chase aircraft, ATC/radar, airspace, telemetry, R/T procedures, pilots and ground crew commenced.
Designed entirely for high speed level flight, Spirit Of Innovation is extremely challenging to operate on the ground. The pilot sits on the left-hand side of what is notionally a two-seat side-by-side cockpit, although it only ever flew with one pilot. Visibility is therefore OK to the left from immediately left of the cowling all the way round back to the left tailplane; close to non-existent at the front and extremely limited to the right.
Even when in the air there is not a great view directly ahead. It is possible to taxi using ‘Spitfire weave’ clearing turns, but it would have required complete turns through 90degrees and would have been an extremely time consuming process.
To assist taxi operations the aircraft therefore has a camera mounted on the bottom of the fuselage just forward of the steerable tailwheel, and on a screen in the cockpit the pilot can see the view forward between the landing gear legs. Also used on the ICE NXT, the camera changes taxi operations from extremely challenging to manageable.
During ground handling runs the pilots – Phill O’Dell and Steve Jones, an Electroflight founding shareholder, former Red Bull air race pilot and recently retired BA 747 captain – had got used to the new situation of going through the detailed checklist to get the aeroplane ready for flight, then having only the noise of the cooling pumps aurally indicate that the aircraft was ready to go.
Steve: “The bizarre thing is when you’re ready to taxi, you turn on the three cooling systems, so you hear three electric pumps turning and of course the prop is stationary. At the beginning you’d call ATC for taxi, and they’d ask you to call back for permission after engine start as the prop wasn’t turning, and you’d have to tell them that you had actually started and were ready to taxi!”
By September the aircraft was ready to fly and on the 15th, up stepped Rolls-Royce Director of Flight Operations Phill O’Dell to earn his test pilot salary on a flight he later described in a very low key manner as ‘interesting’. Rolls-Royce pilot Chris Hadlow launched first in the Extra with Electroflight MD Stjohn Youngman in the front seat, ready to operate as the safety aircraft. Phill lined up on the main runway, gently applied power and became airborne. Hitting the UP button on the electrically driven hydraulic undercarriage, it rapidly became clear that the legs had not travelled fully, and to avoid exceeding gear limiting speed of 150kt it was necessary to reduce power at only 500ft.
The aeroplane is extremely slippery and flying with the gear down requires twice the energy consumption as flying clean, so battery usage went right up. The gear appeared to be hung halfway and didn’t go up so the preplanned response was applied and the gear was selected down, all the time with extremely close attention being paid to the declining state of charge on this prototype aircraft’s first flight.
Climbing to height with the gear down, Phill carried out the test plan of examining stall and performance in the landing configuration. The Extra positioned in close formation underneath to confirm the legs were indeed locked fully down for the landing. With insufficient time to fully examine the aircraft’s handling characteristics and some potential questions about controllability in pitch at low speeds, Phill elected to land flapless and at a higher speed than originally envisaged to avoid getting anywhere near the stall on touchdown.
It became apparent that the ASI might not have been entirely reliable (later identified as a static port issue) and so Chris flew a formation approach with Phill, calling out the speeds on the way down. Touching back down on the runway, the 3km Boscombe runway was fully utilised in length as Phill successfully used the brakes as firmly as was possible without allowing the (very forward CofG) aeroplane to nose over.
“We used a set of brakes – but that’s what they’re there for,” observed Stjohn laconically, “And then people remembered that it was an electric prototype, and all the electric bits had worked fine!”
As with all professional prototype first flights, a huge amount of time had been spent in testing on the ground preparing for the known knowns and the known unknowns, and so despite the startle the whole team were able to calmly and effectively deal with these challenges and move forward.
The undercarriage had worked perfectly in ground tests, but the additional aerodynamic loads prevented the doors from closing properly just at the point where the mechanism had least mechanical advantage. On subsequent flights the gear tended to ‘hang’ on about seven out of 10 occasions, and the pilots used a zero G ‘bunt’ at a safe height to get the gear robustly locked up.
With extensive ground testing of the aircraft throughout the first half of 2021, there were no material modifications to the aircraft in the flight test phase. Changes were focused on a number of software revisions on things such as cooling pump speed, warning parameters and discharge rates etc.
Some 20 flights followed to become familiar with the aircraft’s handling characteristics and expand the handling envelope. Speed was not a consideration at this stage as the team ensured that the aircraft would be safe to fly, obviously on a series of very short flights with specific objectives on each one, and each time ensuring that the aircraft landed with a decent energy reserve. Steve Jones shared what they’d learned about flying the aeroplane.
“Phill and I had a small number of flights each in the conventional NXT in the US, which was tail heavy as opposed to Spirit of Innovation, which is nose heavy because of the weight of the battery pack structure. Pod and I are both on diets to help manage the MAUW.
“It’s an interesting aeroplane to fly, with similar performance to a late model Griffon-engined Spitfire. It doesn’t particularly point where it’s going and the nose is continually wandering left or right, so it’s quite labour intensive and you’re forever feeling for the balance – we even added a slip string on the coaming to help with this. And it’s like that in pitch as well, it has tiny tail surfaces so every time you yaw (or allow it to yaw) the gyroscopic forces on the prop pitches it up or down – or if you pitch, it yaws. So you’re forever putting in these little corrections with your feet and hands to try and keep it on the straight and narrow.
“It has a very nasty stall with a rapid wing drop which you especially want to avoid close to the ground. But it’s all part of the character – I love it! I’ve never flown a Sopwith Camel, but I’d describe this as being like a 300mph Camel!
“The straight and level view forward is not very good and it would be a real issue flying VFR in the UK. Threading your way through airspace around the London TMA is not something you’d consider lightly. It very much helps to operate in a radar environment, and we have every piece of Electronic Conspicuity kit possible in the cockpit.
“But we forgive all that because it goes like the clappers – 550hp feels like a lot. On the time to height record attempt I felt like I was being dragged up by my feet!”
The team liaised with motor providers YASA to determine the optimum rpm. Fascinatingly, the motors have as much torque at 100rpm as they do at 3000rpm – just like a steam train – so they contacted the propeller manufacturer MT and asked them what rpm it would prefer. It felt that 2400rpm was most efficient, and that’s what was mostly used. Experimenting with 2,200rpm gave no material improvement. At 300mph the prop was hugely coarse.
I asked Steve how difficult the take-off process was. “The ‘before take-off’ checklist is very simple, and managing the nose heavy aircraft with very little weight on the tail is critical. Torque is the primary power metric, used in a similar fashion to manifold pressure, and initially is set at 20% to check all three power channels are working as they should, then gently up to 50%, (so the tail is unlikely to lift itself) to get the aircraft charging down the runway.
“Poke the stick forward a little bit to get the tail gently up so you can see where you’re going, then 75% torque for getting off the ground, which is plenty.
“You have to be very cautious, as if you do lift the tail too early you probably wouldn’t be able to stop the pitch moment with back stick, so it all has to be done very gently. The temptation is to lift it off too early, but we need to remember that it has tiny little wings and it’s very heavy, so we just sit there, looking at the view, waiting for it to accelerate to 100kt, then rotate determinedly to get the thing off the ground, and once it’s airborne then it flies very nicely. Then, at a safe height, you bunt to get the gear up. And go very fast!”
The only modification made to the aircraft during the flight tests was the addition of a short piece of door sealant strip or ‘P strip’ to the right-hand side of the rudder. Similar in use to P-strips sometimes used on aerobatic aircraft ailerons, the result was a slight improvement in yaw stability.
The three goals aimed for were for the 3km (to be flown by Phill) and 15km distance records and the time to 3,000m height records (both flown by Steve). The FAI has long lists of rules and regulations for the records, but simply put the height record is measured from brakes off to when the aircraft goes through the specified height. The height equates to 9,843ft and to be on the safe side the team aimed to smash through 10,000ft. Interestingly when this record was actually beaten the aircraft was still climbing at about 3,000ft per minute!
The distance records have more demanding profiles. The 15km record is the average of a run over the prescribed distance in each direction to offset wind effects, but with a mandatory 5km level run-in before each measured leg, resulting in a minimum of 40km in a straight line. Battery endurance meant that Steve would not be able to use 100% torque all the way, so it all became a balancing act in flying as fast as possible, whilst still having enough power to land – for this reason the 15km speed was always going to be slower than the 3km record.
The entire process was very much carried out as a team activity, with Rolls-Royce and Electroflight engineers continually looking to make marginal gains. For an ideal run the motors, inverters and battery would all hit their thermal limit at the same time as running out of energy (with a landing reserve) just as the aircraft crossed the finish line.
With hugely impressive CVs and including automotive engineers at the highest level (with work, for example, in F1 and on the Bloodhound land speed record attempt), Stjohn could not speak highly enough of the joint team that had been assembled. The build standards were of the highest possible quality, and I would have been happy eating dinner off any of the components. Aside from the undercarriage issue, there were no snags from the aeroplane over the whole flight test programme, an absolutely stunning achievement for a prototype.
The actual flying of the turns was more of an issue on the 3km record which required four reversal turns, and to assist in the strategy Max Lamb was part of the team. His experience as strategy manager – nicknamed ‘the aeroplane whisperer’ – for his father and world RBAR champion Nigel, was extremely useful in tracking Phill’s turns and helping to improve times. Initially the expectation had been that close to full throttle would be required from start to finish. One later change introduced involved getting to race start height (actually a ‘perch’ about 800ft above) and then throttling back to allow the battery to cool. This resulted in better efficiency and faster overall times.
Putting the aircraft on the ground after each flight was less than straightforward. Steve explained that the approach and landing are pretty extreme. When they were in the US an experienced NXT pilot shared the best methodology which, précised in his American manner, was to ‘position way too high, way too close on a downwind leg, throw the flaps and the wheels down and dive at the numbers!’ And, apparently, that is pretty much what to do!
“The aircraft is incredibly clean with the gear and flap retracted and incredibly draggy with the gear down,” said Steve.
“I’ve been very tight downwind 2,000ft above the threshold. As the numbers are abeam the wing – there’s no conventional downwind or base – I’ll have the first two stages (of three – ¼, ½ and full) of the split flap down, throw the wheels down and turn. Half way round the turn, last stage of flap and then literally find yourself diving at the numbers at idle power with a high rate of descent, but you can see where you’re going.
“Idle means virtually no battery usage, although the prop will windmill and pretty much maintain 2,400rpm (as unlike a piston engine there is very little drag from the prop), then start to reduce airspeed to move the touchdown point. You add a little torque to cushion the touchdown and gently make sure we don’t flare too high, especially with Boscombe’s wide runway, as the aircraft is delicate and heavy with that nasty stall. We level out very close to the ground and run it on to main wheels as gently as possible. It doesn’t seem to bounce, then it’s a case of just sitting and waiting. The problem is you’re doing at least 110kt and again there’s no drag from the prop, you just have to wait, holding the tail in the air to maintain visibility and protect the tailwheel.
“Visibility is managed by looking out to the left as you need the 3D reference, the camera is not used until the tail is on the ground. The temptation is to brake, but you mustn’t as it’s still going so bloody fast, so plan not to touch the brakes until 60kt after you’ve let the tail onto the ground, and then maybe just take up the slack on the brakes – but definitely don’t brake conventionally else it will nose over. It’s very easy to overheat the brakes and get them red hot as it’s so heavy and fast. Then taxi back using the camera – and home for tea and medals.”
On Tuesday 16 November I was having a coffee with Phill in the Bustard Flying Club crewroom. Breaking records requires operating at the limits and breaking records with experimental technology brings further risk. With decent records established, operations would be wound up at the end of the week but the whole team had a sense that there was more to be gained from the aircraft. A low cloud base hung over the airfield with a depressing forecast for the rest of the week. Phill decided that he’d launch for a weather recce in a CAP 10 to see the weather for himself – and landed for a final run at the 3km record. With a rising sense of expectation, the team swung into action to prepare for the 29th (and what turned out to be) the second last flight of the aircraft.
With some comparatively minor changes to the previous flight profile, Stjohn and I watched Pod launch, get the gear up immediately, slow down to cool the batteries then dive down from 1,200ft for his first 500ft run.
The noise of the prop surprises people expecting a performance similar to an electric car, but it’s still pretty loud. Judged by eye it was tough to tell if this was a faster average time. It seemed smooth and stable, but all of the telemetry and radio interaction was based in a nearby office so we couldn’t really tell how good the runs had been… until Phill taxied in, shut down and opened the canopy with a huge grin on his face. He knew that he’d been close to the record this time, as did Electroflight founder Roger Targett who decided to mimic the aftermath of the Supermarine S6b Schneider trophy winning run in 1931 and gave Phill a piggyback into the office.
Everybody waited while Andy Roberts, Rolls-Royce Head of Flight Test and Team Lead for the flight test phase of the project, went over the figures. In a very low key and almost surreal environment, Andy confirmed that, subject to FAI ratification, the 3km record had been raised to 345.4mph. The announcement was met by what I initially felt was almost a wave of mild indifference rather than cheers and roars. The level heads that stayed calm during unexpected events were similarly level given that breaking the record was largely what the team had been working towards and expected. It was also a bittersweet moment for the project team who now had to move on to the next piece of work.
Steve flew the aircraft on its 30th and final flight the next day, sadly not quite able to better his already impressive 330mph 15km distance record. The aircraft will be used at a variety of exhibitions this year and is pencilled in to be displayed in a national museum, hopefully with the cowls off so that visitors can appreciate the incredible engineering of the Spirit of Innovation team.
What might the team have done differently? With no mechanical failures and the entire electrical propulsion system working exactly as advertised, the project was clearly a massive success. A lighter battery pack would have helped, but as a key structural part of the aircraft I’m not sure any pilot would want to take any shortcuts in that department.
The single biggest improvement to the records would have been location. Pretty much every other aviation record has been set in Reno or New Mexico, with a runway at 5,000ft and a density altitude of 6,000-7,000ft. The team estimated that flying at Reno would have added another 20 to 30mph to the record – which interestingly compares very favourably with the ICE NXT’s record of 415mph. However, the existing electric record was still smashed, and with the patriotic benefit of having been flown from Boscombe Down, the historic and now continued home of UK experimental flight.
The programme has always been about far more than breaking the record, focusing on pushing the technology forward. Spirit of Innovation is a really well-built aircraft demonstrating what electric capability can do. With a focus on high power delivery rather than energy, the aircraft nevertheless managed a flight of 32 minutes at an average speed of 150mph. There is lots of technology that will flow into future urban mobility and hybrid projects, and there is a great battery module which is now marketable for different companies and platforms.
Rob Watson, Rolls-Royce President – Electrical, said, “Staking the claim for the all-electric world-speed record is a fantastic achievement for the ACCEL team and Rolls-Royce. The advanced battery and propulsion technology developed for this programme has exciting applications that will help make ‘jet zero’ a reality.”
This was a real achievement technically for the whole ACCEL team. In the Boscombe officers’ mess on the evening the record was broken, I congratulated Phill who went out of his way to highlight that pilots are just a small part of a much larger team, and he absolutely agreed that every single member had contributed to breaking the record. He paused and then commented, “I’m just glad that I can start eating cake again!”
