Winds of Destruction

by Peter John Hornby Petter-Bowyer

  2. Route to the Target. The route chosen was via Fort Victoria and Chiredzi NDBs and then to the Nuanetsi/Mozambique border which was the initial point (IP) for the bomb run. The aircraft took off at 15-second intervals and routed at Flight levels 100, 105 and 110 to top of descent near Mabalauta. A gradual descent was made to 2,000 feet AGL to the IP maintaining a ground speed of 300 knots.

  3. Run from IP to Target. The IP was overflown at 2,000 feet AGL and six minutes from the target the ground party was warned to light the flare in one minute’s time. At five minutes from target the ground party was instructed to light the manually operated flare and 15 seconds later the fuse was reported burning. The aircraft descended to 1,000 Feet AGL. At two minutes out the manual flare ignited and was observed by all three aircraft. Speed was increased to 350 knots ground speed and all descended to their bombing heights of 300, 400 and 500 feet respectively. As the lead aircraft was about to pass the first flare the second flare (RAMS) was initiated. Final heading corrections were made and all the bombs dropped in the target area. This was reported by c/s 55 and later proven by photography. No sign of any activity was seen by any of the aircraft.

  4. Recovery. The three aircraft recovered to base at FL 250. After landing it was discovered that Green Leader and Green 3 had 50 hang-ups each.

  5. Green 4. Green 4 was heard relaying messages from high altitude during the run-up to target and subsequently as Green 1, 2 and 3 returned to base. He was last heard still at high altitude awaiting the helicopter assault.

  The significance of Randy’s last paragraph will become apparent shortly.

  Schulie and Steven had heard the deep rumble from the open bomb bays then saw the aircraft brightly illuminated by the outer flare as they flashed over their position. Even at two kilometres from the ZANLA base, the roar that came to Schulie and Steven from multiple Alpha bomb explosions impressed the two Scouts enormously.

  Whilst they waited for the flare to stop burning, Schulie heard the Canberras climbing and turning for Salisbury. Then the sound of human screaming and shouting came across the quiet bush from the ZANLA camp. Schulie had to recover both flares for burial well away from Madula Pan. As he did this, he realised that ZANLA had taken a pounding because of the troubled noises that continued to reach him. He was well out of the way by the time the helicopters arrived with the assault force.

  From what they had heard from the high-flying Canberra the lead K-Car crew, carrying assault force commander Captain Richard Pomford, expected to oversee a simple mopping-up operation. Instead they were met by anti-aircraft fire of such intensity that they had to back off and call for the Hunters to strike. At the same time, Ian Donaldson, who was about to turn back to Buffalo Range in the reserve Canberra, was asked to give assistance.

  Ian made a high-rate descent and, for a reason that will never be known for certain, broke through low stratus cloud directly over Malvernia. It must have had something to do with canopy misting following the dive from the freezing conditions at high altitude into warm air with both engine throttles closed. This, coupled with a glaring sunrise ahead of the aircraft flight line, might have been the reason Don did not see the railway line which, had he done so, he would undoubtedly have avoided.

  Rhodesian forces based at Vila Salazar just across the border from Malvernia witnessed extremely heavy anti-aircraft fire with masses of tracer being directed at the Canberra before it rolled sharply and dived into the ground from where a huge fireball was seen to rise. Ian Donaldson, Dave Hawkes and Rob Warraker died instantly.

  The FRELIMO force at Malvernia had obviously been aware of the air attack at Madula Pan and was fully alerted when Ian inadvertently passed over them in perfect range for every one of many heavy and light guns. Some airbursts were seen which suggests RPG rockets may also have been launched at the aircraft.

  Back at Madula Pan, fire from the ground continued to be so intense that the helicopters and Dakota retained their troops, hoping for the Canberra and Hunters to make a difference. They did not know of the Canberra’s fate at the time. However, Selous Scouts Major Bert Sachse, flying as an observer in a Lynx, had watched a number of Hunter strikes before he realised what had gone wrong. His assessment was later confirmed by radio intercepts on FRELIMO’s command channels.

  A large FRELIMO mobile force happened to have spent the night very close to Madula Pan on the road running next to the railway line. These troops were mounting their vehicles to continue their journey when they heard the bomb strike go in on Madula Pan. The whole force turned to investigate and give assistance to their ZANLA comrades who they knew were based at Madula Pan. They arrived before first light to find ZANLA had suffered many casualties and were in a state of panic and disarray. In addition to aiding the wounded, FRELIMO prepared defences for the assault they knew would come with the dawn.

  In the face of this unexpected opposition by a force that was obviously much larger than his own, Richard Pomford called off the assault to avoid unnecessary casualties and Randy returned with three Canberras to deliver eighteen 500-pound bombs onto the target for good measure. He would have much preferred to deliver that load onto Malvernia in response to the downing of one of his Canberras; but airstrikes on any Mozambican town were taboo.

  First radio intercepts reported that ZANLA had suffered six dead and seventy wounded during the night attack. These figures were later updated by FRELIMO with final figures being ten dead and 102 wounded from a force of 120 CTs at Madula Pan. The majority of wounded were described as ‘amputees’, having lost limbs.

  Five months after this, the RLI parachuted into Madula Pan in support of Selous Scouts who were mauling ZANLA and FRELIMO forces close by. It was interesting to learn from them that, though there were no ZANLA in residence at the time, the whole camp area was littered with many human skulls and bones. Presumably these were mainly from FRELIMO men caught in the Hunter and Canberra attacks.

  The Air Force reaction to first casualty reports on the Madula strike was one of great disappointment, particularly by Randy du Rand. We had become so preoccupied with kills that wounding was almost totally discounted in assessing airstrike effectiveness. Randy even advocated reverting to the conventional bombs that we knew did little more than cost us dearly in precious foreign currency.

  I was very distressed by all of this until Group Captain Norman Walsh told Air Marshal McLaren that he believed the Alpha bombs had done a much better job than anyone realised. His opinion was later fully supported by the Army and Special Branch who expressed a different viewpoint from that initially expounded by many Air Force officers. Selous Scouts were the first to expand on this view in writing. Ron Reid-Daly’s contention was that the Alpha bombs had done a much better job by inflicting 90% living casualties with 10% kills than would have been the case if the figures were reversed. Air Staff opinion changed but Randy remained sceptical.

  During Federation days, the Rhodesian and Portuguese Governments exchanged names for two establishments that lay side by side across the common border. The Mozambican town in this picture was named after Lord Malvern (one time Prime Minister of the Federation) and the small village inside Rhodesia (out of sight on left side of the clearly visible borderline) was named after the Portuguese Head of State, General Salazar.

  Note the large empty railway yards that had previously handled heavy traffic moving to and from Lourenço Marques from Rhodesia, Northern Rhodesia and Nyasaland

  The Scouts recognised that dead CTs were either abandoned or buried and forgotten whereas the living wounded presented an unwanted burden by tying up other forces and vehicles in long-range evacuation to rear hospitals. Along the way other CTs and FRELIMO would see these casualties returning from the border, creating a negative impact on morale. Mozambique’s medical facilities had become totally overstretched and CTs with missing limbs and tall war stories had to be cared for. This all created a dilemma for ZANLA who kept war invalids in separate camps away from other cadres.

  The Air Force Commander m
ade it clear that he was delighted with the first Madula Pan strike because it was our best single result to date. It had cost a fraction of the followup air weapons costs with no foreign currency implications. However his greatest concern at that time was to recover the bodies and bomblets from the downed Canberra, providing this could be done without further loss of life.

  For three days the Army made a number of attempts to get to the crash site but vastly superior FRELIMO forces, hellbent on protecting a prize that lay so tantalisingly close to our border, repelled these attempts.

  Then radio intercepts revealed that the bodies had been found, large quantities of ‘ball bombs’ had been collected and, together with Canberra wreckage, all was being loaded on vehicles for transfer to Maputo. In consequence, recovery attempts were called off.

  When we were still developing Alpha bombs, Bev and I decided to stamp the fuses with Chinese hieroglyphics that translated to ‘Made in North Korea’. Whatever the FRELIMO Government made of this when they studied bomblets, later piled next to aircraft wreckage, is not known but national radio and media coverage gave out that the bomblets had been manufactured in the ‘Racist Republic of South Africa’ and that the aircraft had been ‘a gift to the enemy from the British Colonialist Government’.

  Going back to the Alpha Project again—we had produced bomblets that would not detonate sympathetically if one happened to be set off by, say, an enemy bullet. Only a 100G shock could activate a pistol. Both characteristics gave protection against enemy fire and inadvertent mishandling of bomblets. In the case of Ian Donaldson’s crash, the Canberra fuselage had absorbed so much of the shock loading on impact that not a single bomblet had detonated; which is why all 300 ended up in Maputo.

  New Frantans

  MY PROJECT TEAM HAD SUCCEEDED in its most pressing task of providing the Canberras with an effective anti-personnel strike capability. By January 1977 we were already engaged in a number of new projects. These ran concurrently, imposing a huge load on Denzil and Bev who were still heavily involved in the production of Alpha bombs, officially designated Mk2 Fragmentation bombs. In spite of this they had been more than willing to take on new developmental work. Projects Bravo and Delta (pyrotechnics and 37mm Sneb boosters) had been finalised. Next priorities, Projects Echo and Foxtrot, were for new Frantans and high-pressure bombs.

  To provide the Lynx with an effective Frantan, we chose to move away from conventional napalm bomb designs. All those in use in the western world were simply metal tanks, most with small fins designed to pitch the unit nose-down at the moment of release to ensure positive separation from the aircraft. Otherwise, none possessed flight stabiliser fins.

  After release, the tanks behaved in haphazard ways causing them to follow unpredictable trajectories. Tumbling, flying sideways, oscillating and corkscrewing were characteristics that set napalm aside from aimable bombs and made accurate delivery difficult. For instance, if two tanks were released together, with one pitching nose-down and another nose-up, they could land so far apart that one might fall short of target whilst the other passed over.

  When Americans took on a target they dropped four or more napalm bombs from each of a number of aircraft to saturate large areas with flame and intense heat. Absolute accuracy and high costs did not bother them whereas we needed units that would follow a repeatable trajectory to make each unit aimable, accurate and highly effective. This meant we had to produce an aerodynamically shaped unit with low drag characteristics for carriage, but incorporating efficient stabiliser fins to ensure longitudinal stability for clean release and alignment in free flight.

  Metal containers were no good, as we had witnessed on hundreds of occasions. Burster charges coupled with unpredictable case rupture resulted in equally unpredictable distribution of burning napgel. Too often large quantities of the gel remained inside partially burst tanks or sticky blobs of unburned gel lay all over the ground and stuck to vegetation. This was no good at all! I decided we needed tanks that would shatter like glass on impact to free their entire gel contents in a huge fan-like spray of tiny droplets with inter-linking volatile gas. I had learned that the Hunter disposable long-range tanks were constructed from fibres with phenolic resin and that they shattered on impact. We followed this line and produced casings moulded from woven glass fibre and chopped asbestos set in a phenolic resin binder.

  Prototype sixteen-gallon units were made and fitted with Alpha bomb fuses (suitably modified to function at low-impact levels) imbedded in the large pocket of flash-compound that ignited the napgel. From Day One the new Frantans were a great success and, with small modifications, were cleared for operational use on Lynx and Provosts.

  Hunters used imported spun-aluminium fifty-gallon Frantans but these suffered all the limitations we sought to overcome. So I arranged comparative trials between our lowcost sixteen-gallon Frantan and the very costly imported fiftygallon units. The results were astounding. The Hunter pilots were able to deliver the local unit with great accuracy. In itself this was pleasing, but even more satisfying was the fact that the local unit, though only possessing one third of the napgel contents of a fifty-gallon unit, provided consistent coverage of ground that equalled the best of the imported variety. Foreign currency saving was another bonus.

  Golf Project

  IN PROJECT FOXTROT WE ATTEMPTED to produce fuel-air explosive (FAE) bombs, which American military journalists described as having ‘near-nuclear’ effect. One military article was supported by dramatic photographic records of the total destruction of an old US naval destroyer from just one of these FAE bombs. However, destruction of ships was not America’s real interest in FAE. The weapon had been developed to clear large pathways through enemy minefields by detonating hidden mines with excessive over-pressure of ground.

  Ethylene oxide was the medium we employed. There were two reasons for the choice of this liquid gas. Firstly, it explodes with as little as 2% of air inclusion and as much as 95% of air inclusion, whereas most other gases will only detonate within a very narrow gas to air ratio. The second advantage of ethylene oxide is that, when ignited, it produces gas volumes many times greater than any high-speed explosive, such as TNT.

  Each American FAE bomb was dropped at relatively low level and descended to ground on a parachute. A groundsensing device perforated a pressure disc to release the bomb’s pressurised liquid contents at about twenty feet above ground and simultaneously fired flares upwards. The upward and downward flight time of the flares allowed the ethylene oxide gas skirt to widen to around twenty-five metres in radius before the first of the flares contacted the gas skirt setting off a vicious explosion. Lethal over-pressure from a mere five gallons of ethylene oxide dispersed and detonated in this way extended way beyond the edge of the gas skirt.

  Very often the precise positions of CTs firing from dense bush were not known and we had no single weapon that could produce lethal effect over relatively large areas to cater for such situations. FAE seemed to offer a perfect solution to this ongoing problem.

  Considerable time, effort and cost went into Project Echo during which we succeeded in making huge expensive fireballs before, eventually, achieving two terrific detonations. The first of these broke many windowpanes in the Kutanga Range domestic area that was over 500 metres from the blast. What interested us about successful detonations were the sound effects they produced and the fact that they totally stripped vegetation, including substantial trees, up to forty-five metres radius from blast centre. The ground around was pulverised and powdered to a depth of several inches. The sound of each detonation was not a sharp bang, as from TNT, but a loud deepnoted ‘crruuump’ from an explosion, followed immediately by the ‘cruump’ of an implosion.

  Ethylene oxide is a very dangerous substance to store and with Rhodesia being under UN sanctions it was also very expensive and difficult to source. Considering these issues, and realising that weapons that descend on parachutes would be difficult to deliver accurately, even in the lightest of wind condit
ions, we decided to drop the FAE project. Nevertheless, I was still determined to produce high-pressure bombs. Denzil was just as determined and acquired information on the gas-generating properties of every known explosive and combustible liquid compound. His hope was to identify a readily available safe-tohandle explosive that would exhibit similar characteristics to ethylene oxide. When he recommended ANFO we all studied the data before agreeing it exhibited suitable gas-producing properties. This was a pleasing discovery because we could produce ANFO very cheaply and easily.

  Project Golf was initiated by making a direct comparison between an imported 500-pound TNT-filled mediumcapacity bomb and an ANFO-filled 6mm steel casing having equal mass. Both units were mounted vertically on three-foot stands pointing nose down for command detonation from a safe distance. The imported bomb was detonated first. It went off with the usual bright flash, black smoke and a very loud bang with plenty of dust drifting away on the wind. The ANFO bomb was nothing like as impressive to the eye or ear. The explosive flash was nowhere near as bright as the TNT bomb and pasty-grey smoke mingled with dust was drifting off before a deep ‘crrrrump’ was followed immediately by a second ‘crrump’.

  Inspection of the sites showed clearly that we had a winner in ANFO. Loud bangs, such as thunder from lightning, are the product of huge energy releases to atmosphere. In the case of bombs filled with high flame-rate explosives, bright flashes and loud bangs of surface bursts are products of wasted energy following the disintegration of steel casings. When used against buildings, bunkers and other targets where detonation occurs within confined structures, the same energy is highly destructive, but not so in the unconfined conditions of open bush.