RESEARCH AT RAE BEDFORD (1954-1960)

Readers may find it interesting if I weave an outline of my scientific research with some strands of personal life. I will avoid formulae but cite some definitions and references.

I started work at RAE Bedford In October 1954. My first task on my first day was to be briefed on the requirements of the Official Secrets Act and then sign it. Then I was taken down to the 3ftx3ft wind tunnel (the dimensions relate to the size of the working section). This wind tunnel was used to test small models of aircraft or missiles (typically 1/48 or 1/30 scale). It had been supersonic covering the range from M=1.4 to 2.0. However, it was now being provided with alternative slotted working sections, similar to the ones at Brown, covering the transonic Mach number range from 0.7 to 1.2.

I was introduced to the staff (mostly quite young) and given some briefing reports to read. Later in the week came my first challenge: to provide a scheme for a particular model and to design an internal strain gauge balance. These were something of a novelty in 1954. The balances were small enough to fit within the fuselage of a typical model yet strong enough to carry large aerodynamic forces. Typical forces and moments in the old fashioned units then in use were:-

Lift (Normal force) 300lb
Pitching moment 450ft-lb
Drag (axial force) 30lb.

The calibration of these balances required great care and took several weeks. Of course the balances were designed with a factor of safety (3 at transonic speeds and 4 at supersonic speeds) but were only calibrated to the design load.

While this work was in progress, my colleagues at RAE encouraged me to write up and publish the two pieces of unclassified research I had completed in my year at Brown. The first paper¹ was a mathematical solution for a special case of slow viscous flow, when the complicated full equations, reduced to a much simpler form. In the experiment at Brown I had generated a slow viscous flow in glycerine and then compared it to a much faster, less viscous vortex in water, for which no analytic solution was available.

The second paper² was more interesting. This was an exact, analytical solution for an axially symmetric jet with swirl. This was derived from work Professor Squire had published just before I left England. I had to compute some iterative solutions using a hand operated calculator. The effort was worthwhile and both solutions were included in my M Sc Thesis from London University (1955).

On Sundays I worshipped at St Andrew’s, Bedford and served for the parish priest, John Hare. He was pleased when I decided to train as a Reader. He arranged that I should take my three hour examinations (Church History, Church Doctrine, Old and New Testaments) in his vicarage. I was admitted to the Office of a Reader at St Albans Abbey in June 1956 by the Bishop of St Albans. My former vicar in London came up from West Kensington and my parents drove up from Nevendon. I subsequently (1979-82) studied for a Cambridge Diploma in Theology at evening classes in Bedford.

As a relief from my work and my studies for the initial Reader’s examination, I joined the Bedford Amateur Operatic Society. I sang the second tenor parts and went to rehearsals two evenings a week. I was too late for the Pirates of Penzance (1954) but sang in HMS Pinafore (1955) and many subsequent shows. In addition I learnt to play Tennis at the Bradgate Road Club and enjoyed meeting other members. They had a fancy dress dance one Saturday evening. I borrowed an Arab’s flowing robes and went disguised as ‘The Nizam of Hydrostatics’. When my arrival was announced I was not recognised. Then I had the pleasure of explaining exactly where Hydrostatics was located in India.

Early in the 1960’s there was concern that some research scientists lived alone in ivory towers, with little contact with the real world. To rectify this situation some academic scientists were given the chance of a brief attachment to the forces. I applied for an attachment to the Royal Navy, to study what was then my current interest:- Carrier Operations.

I was flown to Malta in a chartered ‘Viscount’ and posted to a large naval airfield at Hal Far (This was beside the former flying boat base at Kalafrana). Unfortunately the fleet was then in Barcelona and the RN ‘Devon’ transport did not have the range to fly from Malta to Barcelona. So my host arranged an interesting local programme for me on the airfield.

On the first day my host was chairing a court martial and I sat with him within an ancient fortress to observe the proceedings. The next day I was taken on a tour of the engineering workshops. I was impressed by the skilled work of the Wrens maintaining helicopters and radar systems.

The airfield had a navigation school, which operated Pembroke aircraft with piston engines. They practised radar homing on any vessels they could find in the sea lanes. The day I flew as a passenger, every ship we found was an oil tanker – proving the continued importance of the Mediterranean. On the Saturday I managed to get a flight to Naples in the Devon. We had good views of Mt. Etna as we crossed Sicily and saw Mt. Vesuvius glowering down on the dense slums of Naples. We had a long wait then from 10.00 to 16.00, the temperature rising steadily in the Devon. (For security reasons we were confined to the aircraft). Then a US aircraft came in with a senior British Officer and we were able to make the return trip to Malta. We were unpopular because we had kept the ground and control staff on duty during a long, hot Saturday when they normally would have been off duty.

I asked if I could make a flight in a Wessex helicopter and was taken up at the end of an ASR practice when a marker buoy had been lost. First we searched for the buoy. Then I asked if we could try an autorotative descent – the emergency drill for an engine failure. The pilot warned the crewmen by the open door to secure themselves. Then he put the engine into fine collective pitch. This made the rotor spin up rapidly as we dropped towards the runway, with the aircraft rotating in the opposite way to the rotor. Then the pilot selected coarse pitch for the rotor and put the engine into ‘gear’. Suddenly we were hovering a foot or two above the runway; it was a powerful demonstration of the emergency descent drill. It had something in common with my earliest spinning test with a Chipmunk monoplane.

The base had several Meteor 7 (twin jet fighters) which were used for target training. I asked if I could have a flight in the rear seat. The answer was yes, but not when target towing – just in case of an accident. The pilot noticed my interest in buffeting and suggested to me that he should demonstrate to me both low speed and high speed buffeting. The low speed buffeting was demonstrated at moderate altitudes and characterised by what I considered a large amplitude, low frequency mode – probably wing first bending. The high speed buffeting was demonstrated at 30,000ft (where I had to breathe oxygen). High speed buffeting was characterised by a small amplitude, high frequency mode – probably wing second bending. After a circuit of Malta at low altitude, (including several tight turns round the famous monument to St Paul’s shipwreck) we landed back at Hal Far.

The armourer came and opened the cockpit canopy and I handed him my parachute. I remarked that “I was pleased that I hadn’t needed to use it”. “So am I sir! Nobody has ever got out alive from the back seat of a Meteor 7.” I was glad I didn’t know that before the flight.

The navy was sorry not to be able to offer me the carrier operations I had requested. As a replacement, I was asked if I wanted a trip in a submarine? I said I would. So I was delivered to the deck of HMS Tabbard at 02.00 one cold morning. The boat sailed out into the Grand Harbour to begin the exercise. Once through the harbour entrance, the sea got very rough, with huge waves drenching the conning tower. I knew we would be diving soon so I retired to the wardroom and the warmth and comfort of a bunk. The steward asked me how I liked my breakfast eggs done. (I think my 3½ minute egg at home needed fewer minutes in the submarine because of the increased pressure). The officers settled down on their bunks until a wake-up call sounded. My recollection is that as the officers pulled on their clothes they swore profusely, without repetition. We had breakfast and then the exercise started.

The object of the exercise was to attack and ‘sink’ the cruiser Lion. We were getting close, with good views of the target through the periscope, when suddenly our sonar signals were picked up by two destroyers. We had to crash dive and then furtively sneak away from the scene. We were lucky to escape with the boat….

After our escape, we resumed patrol under the surface and I was taken on a tour of the submarine. My recollection is of a cheerful and enthusiastic crew, but I didn’t like the oil, dirt and darkness. The submariners just said “but the pay is good, Sir”.

When we surfaced, I went on the conning tower and enjoyed the sunshine. The skipper asked me if I would like to stay for a drink in the wardroom and to have a chat. I said “Yes please, but I have a van coming for me at 13.00 to take me back to lunch at the Mess.” “Oh, we can easily alter that by sending a radio signal”. So the famous signal was sent from HMS Tabbard – “Mabey late (15.00). Save lunch.” - I often wondered how the Russian Cypher specialists interpreted our signal. But we did part good friends; a fitting close to an interesting attachment. I thanked them that my attachment had included a voyage in their submarine, as well as the high altitude flight in the Meteor. They gave me good wishes for my marriage, now only three months away.


A GLIMPSE OF FAMILY LIFE

In 1957 the Bedford Operatic Society was joined by a very attractive young lady who helped with the make up of principals and chorus. She sang the alto line in Ruddigore in December 1958 and was selected to demonstrate how to dance the Gavotte with the producer. That crucial moment (when she caught my eye) was recalled years later in a poem ‘Put the Clock Back.’ Gill Warren was a graduate in General Sciences from Kings College, London. She worked as an Information Officer at the Unilever Research laboratory at Colmworth House, Sharnbrook (North Bedfordshire).

 

(a) Gill Warren (b) Gill Mabey
(Langdale – June 1959) (North Bedfordshire – Autumn 1960)

Fig 3 Gill


Gill and I married on 24th September 1960 at Burgh Heath (Surrey) and moved into a new, semi-detached house in Bromham. I left St Andrew’s, Bedford and became reader at St Owen’s, Bromham. Our marriage was very happy and blessed with two daughters. Deborah Jane was born in January 1963 (in the depths of a bitter winter) and Katherine Anne in July 1966 (in the middle of a hot summer). I subsequently wrote sonnets to celebrate each birth. Deborah’s birthday (23 January 1963) and Katherine’s birthday (17 June 1966).

With young children in the family, we had to restrict our participation in the Operatic Society, which took much time (with rehearsals two evenings a week). The normal pattern was that Gill and I would alternate from year to year. In addition to the Operatic Society, Gill developed a wide range of other activities; dressmaking, dress design, Bedfordshire Lace work, Bedfordshire Rush work, flower arranging and local history. Gill was a keen gardener and when my Mother died (March 1975) we moved to a detached house in Bromham with a much larger garden (October 1975).

I had been improving my swimming by taking stroke improvement and life saving courses with the Modernian Swimming Club (MSC). This was useful because both Deborah and Katherine were becoming keen swimmers. I qualified as an ASA swimming teacher in January 1976 after a course arranged by Miss Helen Elkington*. I then taught swimming both for the MSC and Robinson pool in Bedford.

Deborah married Mark Everett in April 1986 and Katherine married Clive Lewis in July 1988. Gill made all the beautiful wedding dresses and flower arrangements. When presenting papers at scientific conferences or meetings, I took Gill with me whenever I could. Together we visited Stockholm, Amsterdam, Brussels, Gottingen, Paris, and even Madrid (October 1989).

Sadly, in April 1989 Gill developed a rare form of Cancer which did not respond to treatment. During her long illness I found myself writing many poems, some of which she read. Just before her last session of chemotherapy we stayed four nights in a small pub in the village of Oddington in the Cotswolds (recommended by our younger daughter, Katherine.) One day we visited the parish Church at Fairford to see the magnificent pre-Reformation stained glass windows, which had escaped destruction by the Puritans. As we left, Gill spotted a small tombstone. “Who’s that small stone for?” I walked away from the path to examine it. “It’s for Tiddles, The Church cat, 1963-1980” Gill was very fond of cats and I resolved then to write a poem about Tiddles for her (although I dislike cats). That evening at the pub we listened to a radio programme about Thomas Hardy’s poetry. One poem was about the burial of his cat under a tree in his garden. I went to sleep with both cats in mind.

I woke up in the early hours with the first line of “Elegy for Tiddles”. I couldn’t put the light on, so I had to compose the whole poem and memorize it. I wrote it down at the first opportunity (In mid morning) and gave it to Gill to read (she did not like my recitations). She read it, and gave it back to me with the comment “Hm! Not bad!” which was praise indeed.

I wrote a poem “A Baptismal Hymn” (with an original tune) for the Christening of Deborah’s first child, Christopher in June 1991 at St Mary’s, Oakley. Gill was in a wheel chair for the service, but at the end of the hymn she turned to me and said “I did like your hymn!”


*Helen Elkington was the coach for the British ladies syncro – swimming team for many years.

Gill bore my poetry and her illness cheerfully and died at the Sue Ryder Hospice at St John Moggerhanger on 27 November 1991. (Gill had been a back-up helper at the Hospice for several years as she had her own car and thus could get in at short notice). I published my first collection of poetry in aid of St John’s, “St John’s Wort”, and dedicated to her memory in 1999. I thought that after her death the poetry would stop coming – but it didn’t. I published my second collection (also in aid of St John’s Moggerhanger). “Wait for the Wind”, dedicated to the memory of the Wright Brothers. The Wright Brothers first flew on 17 December 1903 and I have long admired their achievement. Later I made many walking holidays with the Ramblers and wrote rambling poems. These remain unpublished.


SOME SELECTED UNCLASSIFIED RESEARCH AT RAE BEDFORD (1960-1991)

Despite the difficulties, I offer the reader seven* crucial examples of my unclassified research. Much of my early work on the structural response (buffeting) of aircraft due to the aerodynamic excitation caused by separated flows (buffet) was summarized in my survey paper “Beyond the Buffet Boundary”. The Royal Aeronautical Society awarded me my first Busk prize in 1973 for this paper³.


A COOLER PROBLEM

A new cooler installed in the maximum section of the RAE 3ft x 3ft tunnel caused a severe problem under test. As the wind speed increased, the tunnel started to roar and to vibrate with large, low frequency amplitudes. Loose bolts on the roof of the tunnel were bounced up and down and then fell over the side, endangering the life of anyone walking underneath. I thought the problem over in my bath (truly!) on a Saturday evening. Could the problem be a transverse acoustic resonance excited by vortex shedding from the circular tubes within the cooler? Then I remembered a reference I had found about a similar problem. First thing on the following Monday morning I traced the reference. Sure enough, some American designers had made a similar mistake.

Our designers had put the new tubes horizontal for ease of cleaning. Thus the vortices being shed were exciting vertical standing waves. To verify this hypothesis, I had to stand in the tunnel just upstream of the cooler with a hand-held pressure transducer. With the tunnel running, I measured two vital unsteady parameters:

(1) The linear variation of vortex shedding frequency with velocity and
(2) The wave form and frequency of the standing wave.

With the tentative diagnosis confirmed, an immediate cure was demanded. I suggested that we might be able to tackle both problems simultaneously. We had very skilled carpenters at RAE Bedford. Suppose thin plywood splitters could be pushed between the rows of tubes? Then these splitters would impose a streamwise velocity gradient on every tube, making them shed a lower level of excitation, albeit over a wider range of frequency. In addition, insertion of only two splitter plates would raise the resonance frequency by a factor of 3, uncoupling the resonance frequency from the vortex shedding frequency. A small section of the cooler was modified quickly and a complete cure demonstrated for that section. Then the whole cooler was treated and the aerodynamic noise and vibration were eliminated4.


*In Holy Scripture the number seven symbolises completeness!

RESONANCE FREQUENCIES IN VENTILATED WIND TUNNELS

Having observed the severity of the transverse resonances in the new cooler, subsequently I used circular cylinders to excite transverse resonances in a small transonic tunnel. (In most transonic tunnels the walls are ventilated by slots or perforations). The cylinders excited strong resonances but revealed that the accepted theory for the resonance frequencies in ventilated tunnels was seriously in error. This was primarily because the original theory took no account of the depth of the plenum chamber around the ventilated working section.

I confirmed this major anomaly by rapid tests of cardboard ventilated working sections with the air at rest and hence M = O. At this Mach number the difference in frequency according to the new approximate theory I had developed was largest. I verified the new theory at subsonic and transonic speeds in two wind tunnels with slotted or perforated walls5. The Mach number is the ratio of the free stream velocity, U, to the local velocity of sound, a, M = U/a.

SCALE EFFECTS AND CRYOGENIC WIND TUNNELS

Our 3ft x 3ft transonic tunnel was used in attempts to evaluate “scale effects” on our small models. Scale effects are caused by the viscosity of the fluid (usually air) and are determined by a non-dimensional Reynolds number.

R = Ul / V¯
Where U = free stream velocity
l = relevant length (often a wing chord)
V¯= kinematic viscosity

Probably the most well known scale effects are those that determine the drag of an aircraft. Our small models gave small Reynolds numbers (typically only about 1.5 x 106) compared to full-scale values of about 15 x 106 to 45 x 106. Hence these small models were unlikely to represent even the drag of the aircraft, quite apart from representing the full-scale buffet onset boundaries. How could higher Reynolds numbers be achieved?

My friend Bob Kilgore of NASA Langley had suggested that tests in a cryogenic tunnel (with gaseous nitrogen at a static temperature of 100к should be preferred to tests in air at 300к). This difference in temperature would reduce V¯ by a factor of 5 and hence increase R by a factor of 5. This increase in Reynolds number would take even our small, 3ft tunnel models from 1.5 x 106 to 7.5 x 106, into the range where scale effects were much smaller. I discussed this problem with Bob Kilgore, who remarked that “Old Dame Nature” always works for cryogenic wind tunnels. Subsequently I was invited to speak on two controversial topics at NASA Langley.

The first topic was the design of transonic tunnels with low levels of flow unsteadiness6. The second topic was on the many advantages of cryogenic tunnels7.




I introduced this talk with a slide based on a picture from Deborah’s storybook.

 

Fig.4 Advantages of Cryogenic tunnels
for aeroelastic tests

The little girl was wearing an apron that acknowledged the higher Reynolds numbers generated by cryogenic flows. In addition, her Grandmother (Bob Kilgore’s Old Dame Nature!) was giving her two other presents. In a cryogenic tunnel one single model can represent the Full Flight Flutter envelope. In addition, these models would have to be heavy (and hence stronger) to satisfy the requirements of aero elasticity. To verify these concepts I suggested tests on the 65° delta and unswept models shown. NASA took up this suggestion and subsequently invited me to help report the measurements8.

In 1975 NASA made the decision to build a large cryogenic transonic tunnel at Langley (the National Transonic Facility – NTF). Subsequently a cryogenic European Transonic Wind tunnel (ETW) was built near Cologne.


SOME AERODYNAMIC RESEARCH FOR CONCORDE

The RAE was deeply involved in the development of Concorde. After many static stability tests, an ogee platform was adopted to reduce the movement of the wing centre of pressure from subsonic to supersonic speeds.

While testing two different slender wing models, I evaluated the large scale effects due to the bands of distributed surface roughness used to fix the transition of the boundary layer*. These measurements established roughness drag penalties that were used as guides in subsequent tests and thus helped to predict the correct full-scale lift/drag ratios9.



*The boundary layer is the narrow region between the surface of the model (where viscosity holds the fluid at rest) and the external region where the velocity is close to the free-stream. Scale effects are generally small with attached boundary layers but much larger with separated boundary layers.
Initially it was considered that slender wings would not suffer from buffeting, because they had a well organised vortex flow. However, wing-root strain gauges applied to all slender models tested in the RAE 3ft tunnel showed two distinct types of buffeting:

Light buffeting (when the wing vortices formed at an angle of incidence
of about 6°)
and Heavy buffeting (when the wing vortices burst at an angle of incidence
of about 22°)

Hence, at full-scale light buffeting occurred on every flight – more noticeably when landing than when taking off. Heavy buffeting was excluded by limits on the control system.

It proved difficult to quantify the level of the light buffeting, even when using a 1/55 scale aeroelastic flutter model with an ogee planform (Fig.5). This model had mode shapes and frequencies much more representative of the aircraft structure than the ordinary wind tunnel models.10 On the aircraft when landing the wing tip motion was clearly visible at a frequency of about 5Hz. Buffeting at this frequency could be felt in the rear fuselage.
 

Fig.5 Aeroelastic flutter model

I flew in Concorde from Toronto to Heathrow in May 1997. Like all the other passengers, after our champagne lunch we were only allowed 30 seconds in the cockpit. The first officer was flying the aircraft from the starboard seat. I showed the Captain my card and said that as a young man I had been involved in the development of Concorde. He made no response. So I then asked him “Had he lost any rudders lately?” (I had been invited to comment on a series of rudder failures). He replied “No! But its nice to get someone like you in the cockpit from time to time!”

I considered Concorde a visionary concept and a unique technical achievement. Hence I watched with sadness the retirement flights of 3 Concordes into Heathrow in the autumn of 2003.


INFRA-SOUND PROBLEM

In 1978 my group was loaned a temporary hut which was close to the RAE 3ft wind tunnel. Scientists working in the hut quickly experienced loss of concentration and a general feeling of tenseness when the tunnel auxiliary reciprocating compressors were running. This was tentatively attributed to high levels of infra-sound from the compressors. Infra-sound is sound at a frequency below the normal audible range (about 15Hz).

It was therefore suggested that the hut should be moved. The implementation of this suggestion would have involved disturbance and considerable expense. Concerned for the health of my staff, I attempted a brief examination with the assistance of the safety officer and a Bruel and Kjaer sound meter.

These results were reported fully elsewhere.11 The crucial finding was that at the discharge point from the compressors (two normal holes on a north facing vertical wall) the sound level was about 130dB (the maximum level of the sound meter!).
At such a low frequency a conventional exhaust muffler would have been very large. Instead two 90° elbows were attached to the holes and set at angles of about 30° and 50° to the horizontal to diffuse the sound and to direct it up and away from the roof. Short lengths of plastic pipe were then added to ensure that the sound was carried away from the North-facing wall, which had acted as a baffle plate. The pipes terminated just beyond a 90° angle formed in the East wall of the tunnel, so that any sound propagating in the opposite direction to the discharge would also be reflected away from the building.

These measures greatly reduced the levels of airborne sound. On the roof close to the discharge point the sound level fell from 130dB to 90dB. In the noisy room of the library the sound fell from 86dB to 78dB. (This level is acceptable for conversation). In the hut the sound level fell from 84dB to 78dB. However, the hut was only lightly constructed and the level of ground-borne sound was about 76dB. Scientists working in the hut found it much more comfortable and the hut did not have to be moved.

This installation operated for many years with a dangerous level of airborne sound. Many other compressors within industry probably discharge under similar conditions and could be improved in a similar way, with advantage taken of the shape of the building.

BLAST FROM MOVING GUNS

The next topic selected was unusual and required a major team effort within the RAE. A research scientist, Frank Smith, had developed a unified theory to predict how gun blast might scale with different parameters. The first parameter varied was the physical size of the projectile (say from a rifle bullet to a 16inch naval shell). The second parameter was the altitude of the projectile (say from sea level to 40,000ft). The third parameter was the speed of the gun (say from M=0 to M=1.8).
For the first parameter Smith had established a correlation between his own recent test on a 7.62mm NATO rifle (at M=0 at sea level) and historic US naval gun blast measurements made in the 1920s. Smith’s initial test suggested that the same rifle would work well in the RAE 3ft tunnel. In addition the 3ft tunnel could vary the other two parameters. The static pressure could be varied from about 0.2 bar to 1.0 bar and the Mach number varied from M=0 to M=1.8. The tunnel thus appeared suitable for this demanding project – subject to one crucial caveat. How were we to stop the bullets before they ran through the screens and the coolers?

Smith referred me to the Weapons Department at Farnborough – who could also supply a NATO rifle and an armourer to fire it. Weapons Department recommended designing a strong steel catch box and packing this with telephone directories (to absorb the energy of the bullets). We designed a suitable catch box for the settling chamber upstream of the tunnel working section (Figs.6a&b) and quickly showed that if worked well for M=0 at atmospheric pressure.

 

(c) Typical blast waves

Fig. 6 BLAST FROM MOVING GUNS
The results of these experiments were reported elsewhere.12 They suggested that Smith’s theory gave an accurate method for predicting the arrival times of blast waves upstream of moving guns and a fair indication of the blast loads on adjacent surfaces. However, downstream of the muzzle and close to the gun there were some discrepancies due to the local flow (which was not included in the theory).

While I was busy co-ordinating these trials my vicar suggested that as a Reader, I might like to preach a sermon about God and my work. I declined politely, explaining that “God and my Gun Blast” might make an unfortunate newspaper headline.


HERCULES TANKER PROBLEM

One Friday afternoon in June 1982 I received a phone call in my office alerting me to a problem on the Hercules aircraft. Some RAF Hercules were being converted for flight refuelling by Marshals of Cambridge. The objective was to establish an “air bridge” between Ascension Island and the Falklands (see Fig prepared to introduce an oral presentation). Unfortunately, with the original modification there was elevator and rear fuselage buffeting and the refuelling drogue was not deploying safely. Marshalls had proposed an aerodynamic “fix”, which would be test flown that evening.

 Click here to return to 'Dennis Mabey's Autobiography' Main Page