The Aerodynamics of the MiG 21 accidents

Prodyut Das M.Tech. PGCGM, M.Ae.S.I.
Former Professor Mechanical Engineering IIT Kanpur

Are MiG-21 accidents really out of control as the press says it is? or is there a pattern behind the accident rate? alt

A fighter pilot flies a wing and an Engine. The qualities of the wing like the wing loading, span loading, aspect ratio, section profile of the wing and the qualities of the engine such as power loading and response time determines the flying qualities of a Fighter.

Fighter flying is one of the most hazardous occupations known to man. It is hazardous because the very high speeds, low level flight over inhabited areas in an airspace often attractive to large birds means that both the reaction times and the options available to a pilot in an emergency are often dangerously limited.

This spate of MiG 21 accidents caused much anguish and was widely written about. This paper is an analysis the technical and statistical aspects of the accidents and has some suggestions of future importance.

The MiG 21: a Technical appreciation

It is significant that for 50 years the MiG Design Bureau were at the forefront of fighter design and yet they never used anything but "yesterday's" proven technology to set and maintain the pace ( Note 1).

There is a lesson in that for Indian aircraft designers..

Coming from such a distinguished pedigree it is not surprising that the MiG 21 was remarkable. It combined low cost proven technology with brilliantly innovative and insightful application of the physics to produce at lowest cost the solution of a high altitude bomber interceptor.( Note 2 ) At the time of its induction into the IAF the MiG with its auto stabilization, radio altimeter, fully duplicated controls and general reliability introduced new standards of safety and reliability .Pilots converting to the MiG 21 were universal in their praise for its ease of piloting and safety and the reliability and the functionality its systems. 'Safe as a bullock cart" was how the aeroplane was described in the late '70s.

The Aerodynamics of the MiG 21 in low level flight.  

Unfortunately all fighters are designed primarily for air superiority but end up in the more hazardous low level close support role.

This was also the case with the MiG 21.From the 1980s the MiG switched to the close support role. New upgrades to make the type suitable for close support also meant a steady increase in weight. The aircraft became more sluggish and unwieldy particularly during the landing and take off and in circuit where the aerodynamic control forces decline as a square of the flight velocity but the inertias remain the same. The weight increase affected the wing, power and span loadings (please refer toTable A- for the MIG the figures on the top of each box are for the FL those below are for the Bis).

The span loading increases is a good indication of how much more angle of attack has to be generated at a given speed to maintain height. Increase in the angle of attack in turn means more power to stay aloft.

The wing loading increase shows how much more speed has to be increased to maintain level flight. A doubling of wing loading would mean a 40% increase in flight speed. This also means a doubling of the power required.

The power loading indicates how much power is available to accelerate the aeroplane should the airspeed fall too low. I have computed the figure for max dry thrust as in a crisis there would not be those few seconds available for the afterburner to kick in before the aircraft impacted.

A special mention must be made of the low aspect ratio of the MiG 21. The low aspect ratio makes the aircraft "alpha "sensitive. The CL /Cd curve becomes unfavourable in low aspect ratio wings. In other word unless the pilot gets the angle of attack right he may see a very great increase in the drag of the aircraft without any corresponding increase in lift. His total energy would decay preventing the aircraft from accelerating. Translated into reality it means one of the following scenarios: During take off "over rotation" -too much nose up-would mean poor acceleration due to high induced drag and failure to lift off with the aircraft running into the overshoot area at high speed.

During landing the misjudged alpha would increase the induced drag causing the aircraft to slow down, lose lift and hit the ground before reaching the touch down area.

During turn into the finals ( or during low level aerobatics) the aircraft is pulling more 'G"s with corresponding increase in induced drag slowing down the aircraft which is already side slipping because of the turn and losing height over ground. A combination of side slip during a turn with high induced drag reducing speed caused unforeseen height loss and a "controlled flight into terrain".

Very many of the MiG 21s lost were in these three regimes of flight. Even in civil airliners most accidents occur during these three phases but:

1) The alpha sensitivity of the MiG21 ,because of the low aspect ratio of 2.2, requires much more precision than the same maneuver when executed in an aircraft with a aspect ratio of 5.6 as in a basic trainer.

2) The continual, if inevitable, weight increase in the MiG 21 meant that the approach speed in the later marks had to be made at a higher and higher speed. This reduced the amount of surplus power available to accelerate away from a "coffin corner "situation". In India the hot weather meant the engine was producing about 12%less thrust and the wing was producing about 12% less lift to begin with.

3) In case of an emergency, to gain height, the pilot in a Hunter or a Kiran would open up the throttle and pull back the stick- things which are instinctive even in a rookie pilot. Ina MiG the pilot has to push the stick forward, build up his energy and then after a delay of several seconds, pull back the stick to climb away. He may simply not have the time when flying close to the ground.

4) The CK ejection seat, one of the best for high speed high altitude ejection simply was not good enough for low level by modern standards. One of the clever features of the CK seat was that as the seat left the cockpit the canopy- which was hinged to the front of the windshield in the FL - attached itself to the top of the ejection seat and rotated itself until it covered the entire front of the ejection seat- thus giving unparalled blast protection when ejecting at supersonic speed. I remember a Martin Baker engineer getting very interested in how the thing worked. I had seen the seat but he had not! Unfortunately I was not able to help him. The semi -encapsulation feature delayed ejection in that it took too long to get rid of the canopy after clearing the aircraft and this delayed clear release and deployment of the Parachute. The 300 kmph, 100 meters minimum parameters meant that many low level ejections were unsuccessful.

The span loading, wing loading the power loading and the aspect ratio of a series of aircraft flown by the IAF is tabulated at Table A.

Accident rates in supersonic fighters

The accident rates in supersonic fighters of the same generation as the MiG 21 makes for relevant comparisons.


The German Luftwaffe flew about 950 F 104s from about 1960 to 1987 and lost about 292 of them during the same period. Average loss rates were thus about 11 per annum though the peak loss rate was 28 aircraft in 1965 and about the same in 66. The ejection seat of the F104 was even worse that the MiG 21 for low altitude flying. The Germans corrected that by switching over to the Martin Baker GQ 7 seat sometime in the mid sixties.

The Canadians lost half of their fleet of 200 CF 104s during a similar period of service. Training was admittedly a problem with the German Luftwaffe which was barely ten years old at the time of the induction of the F104 but the same could hardly be true about the Canadians.

The true master of the F 104 was the Spanish Ejercito d Aire who never lost a single Starfighter in seven years of service. People said it was due to fine weather over Spain! This usually made the Spanish AF indignant! The British RAF lost over a hundred of their 297 Lightnings in about 25 years of service- a number of them to engine fires which was probably due to a flaw in the detail design. Our MiG losses have been at a much lower rate.

Flying fighters is a hazardous business and continual stress and training on flight safety and discipline without killing the spirit of the Fighter Pilot is a difficult and skilled art.

Interestingly the Pakistan Air Fore lost 23 F-7s ( Mig 21 equivalents ) in 10 years which approximates the IAFs loss rates, given their smaller fleet size. It is stated that the PAF possibly reports only those crashes which are in populated areas. In fact the PAF is extremely touchy about anything that shows it in a bad light and it would possibly be that their actual crash rate is higher than the IAFs MiG 21s.

About their image consciousness I remember in the mid 80s there was a review on the PAF in Air International and there were these enormously dirty and dusty MiG 19s with chipped yellow and red paint and oil streaked fuselages and generally looking very neglected though the MiG 19s were right there on the flight line. Someone obviously got (and deserved ! ) a "rocket" because since then I have never seen a PAF aircraft in any magazine that did not look as if it has just come out of the paint booth.

One notes that there are 4 mid air collisions in the 23 PAF losses reported- the MIGs outward view was never its strong selling point. I feel the mid air collisions were more for this type. A switch to a glass cockpit and bulged hood ( to raise the pilot's eye line) may be useful at least in improving forward view.. Another reason could be due to the very small wing span in comparison to the fuselage length. In formation flying the aircraft would be that much closer together. In a turn, with the poorer visibility sometimes the collision was inevitable.

What should be the accident rates? Ideally zero.

However this is not possible in a profession where getting back with a tale to tell or making a hole in the ground can depend on decisions made with a difference of hundredth of a second or a few meters difference in position or height.

The Western world considers acceptable an accident rate of 1 in 10,000 hours as an 'acceptable". If we can accept this figures at a face value then a simple set of "expected number of accidents " could be created by assuming the number of squadrons and assigning a certain percentage of availability and a certain number of flying hours. This would work out to about 7 aircraft per annum in the 80s and thereafter. The fact that in the tropics the aircraft is flying in a non ISA atmosphere means that the engine thrust and the lift available is lower than available to a European pilot. In addition the flying environment- the ratio of open to densely populated areas, the number and size of birds at low level can alter the accident rate in spite of identical standards of training and maintenance.

Table A .

Comparative parameters of IAF aircraft and also the Chinese FC 17

Type Power Loading N/kg Wing loading Kg/sq.M Span Loading Kg/M Aspect ratio Aerofoil Thickness Pitch Inertia Kg.m*mX10e6
Harvard 2.75 93.3 172 6.96 15% (?)


Vampire 3.8 159.8 355 5.51 12% ( ?)


HJT 16 3.71 157 280 6.025 15% 0.309
HJT 36 IJT 4.6 183 324 5.56 15% ( ?)


Hunter 6.0 228 722 3.24 8 % 1.16
MiG 21 5.6/8.2
(with a/b)
279 334 897 1076 2.2
all models
6 % 1.18(FL)
Su 7 7.3/10.5 274 1019 2.65 10.9 % 2.2513
F 104 6.94/10.41 417 1142 2.44 3.5% 2.12
HF 24 Marut 6 256 799 2.89 6 % 1.387
Hawk 5.4 254 458 5.2 10.9 % 0.5359
FC 17 6.6/ 11 319 816 3.52 6 %




1. The above table gives a good insight to the problem. In the earlier days the pilot progressed through the HT 2 to the Harvard and Vampire to the Hunter. The critical parameters of Wing, power and span loadings and the aspect ratio progressed gently and even at the then top of the line aircraft the Hunter the aspect ratio was relatively modest.

As long as the Hunter was there in service a kind of de facto advanced trainer available to stream in the pilots to the tricks of "power flying". The pitch inertia figures are approximate but shows why pilots translating from the responsive Kiran would have found the MiG 21 slow to respond in pulling up or down. Criticism of the Marut and the Su-7 in terms of slow pull out after a close support run can also be found in their relatively high pitch inertias.

With the phasing out of the Hunter the pilots translated directly from the HJT 16 Kiran to the MiG 21 where the figures increased several times. The pilot had to be alert to keeping his energy levels within bounds. Whilst this was also generally managed, it meant, in a case of a mistake, the pilot was skating on thin ice.

2. The accidents were evenly distributed between seniors and rookies. Of the100 cases of accidents where the pilot is identified by name we have 36 accidents where the pilots were Squadron Leaders and above and 24 were of the level of Flight Lieutenants. Only 40 of the accident cases were below these ranks and one could ascribe inadequate experience as a cause. To note however is that 60 % of the accidents, in a sample of 100, involved senior pilots.

3 Of the 164 losses between 1962-2004 that is recorded in the Warbird of India records the main categories were:

Cause Nos Lost
Mid Air Collisions 10
Bird Strikes 10
Take Off or Landing Phase 29
Combat Related 11
No details in Public Domain rest

The numbers may appear inadequate as a statistical sample. However one has seen or dealt with an infinite sample (so beloved of statisticians!) but in practice the laws work as well for a sample of 50 as for infinity! Even if we take the first three cases ( leaving the combat losses out) it is statistically reliable. It clearly shows that accidents where that low aspect ratio was a factor ( i.e TO & landing ) dominate with 29 cases out of 49 i.e greater than 60%.

Is the general accident level too high? According to the MOD the IAF lost a total of 315 MiG 21 were lost from all causes in 40 years. Taking out the combat losses this is less than 30% of the MiG fleet in 40 years. Compared to the Canadian losses of 50% and the German Losses of 292 out of 915 F104s in a much shorter period of operation and the fact the percentage of Lightning losses for the RAF were just as high would indicate that losses were not unusual.

The figures from Warbirds of India are likely to be incomplete, especially in the pre 90s figures. However there are some statements made by the MoD that might help us. Using the figures the Warbird's loss records can be modified as follows.

The loss as recorded in the Warbirds site:

Financial Year No of Aircraft Lost

The figures would indicate that the losses from 98 to 01 are "out of control" But if we modify the figures by using some figures later given by Mr. Pillarisetti (Source: the Warbird thread quoted by Mr. Pillarisetti and the statement by the MOD about the MiG 21s lost during the period we can construct a closer model of the actual aircraft losses .

In this model the numbers "missing " in the Warbirds site with respect to the MOD statement are equally distributed into those years of the Warbirds site which showed a lower number. The actual crash figures would thus be:

Financial Year No of Aircraft Lost

(Thanks to Jagan Pillarisetti for this table)

Looking at the revised figures we can say that the losses whilst regrettable do not reflect any sudden decrease in quality.

There are many press reports that said that the dip in 1997-98, where only 7 were lost could be because of ACM Sareen's measures of reduced flying efforts, which was the butt of much criticism. 01-02 drop also indicates some kind of "interference". The view was that the high value for 98-99 was due to Kargil which meant extra flying and the state or readiness thereafter.

The sudden decline of crash rate après 2003, is statistically speaking, "out of control" and indicates the presence of a new factor which powerfully reduced the trend.

The reason for the crash rate as also the sudden decline in recent times can only be conjectured. The popular press mentioned everything from inadequate training (which was probably not true) to spurious engine parts from ex- Soviet Republics to combustion cans losing their enameling. None of this can be verified.

According to Western rates the loss rates should have been around 5to 7 aircraft a year but one must remember that they do not have to contend with large birds and high runway temperatures. Pilot attitudes may play a major part in accidents. "Disciplining" pilots may reduce the accident rates but break the spirit of the fighter pilot which is counterproductive. Somehow the RAF manages to maintain a balance in an understated way and one expects the IAF has its own methods.

Thoughts on Future training equipment

4. The induction of the Hawks will be warmly welcomed but judging by the parameters in table A and the above para 2. It would be unwise to expect any dramatic reduction in accident rates because of the introduction of the Hawks per se.

It is interesting to conjecture whether HAL should prepare a few IJT with a thin, low aspect ratio wing as a Mk 2 IJT. This should have a 6% wing and an aspect ratio of around 3.2 and a smaller area which would improve the thrust to weight ratio and push up the wing and span loadings but keep the same basic systems. Pilots qualifying on the IJT could then easily change to the Mk2 to extend their training envelope by flying a relatively more "snappy yet similar" machine before proceeding to the AJT.

The LCA has an unusually low aspect ratio of 1.9. It will inevitably put on weight in mid life. Unless it has been tamed by the FBW software, the LCA , will be requiring much careful handling at low speed low level flight. It is also a single engine machine. Will it repeat the MiG experience? May be but the loss of life will be less as it has a very good ejection seat.

5 The importance of having the best ejection seat possible cannot be overstated. It is noteworthy that the Pakistan Air Force retrofitted their MiG 19s with the Martin Baker MK 10. Such a seat in the MiG 21would have saved many of the 70 pilots killed. Wg.Cdr Gautam, MVC and Bar who died "dead sticking" a MiG 21 FL during take off at Lohegaon is one name, of the many, who come to mind.

6 HAL made many valiant attempts to revive the HF 24 Marut. The table does show what a good potential it has as an advanced trainer. With the same Saturn AL 55 engine possibly with an afterburner and systems aggregates as the HJT 36 and the avionics of the MiG Bison for commonality and a modest use of composites in non critical parts it is almost spot on advanced trainer cum long range strike aircraft. It has a lot of room for modern avionics. .A time to first flight of 36 months and an IOC of 54 months should be achievable. How about involving the private sector? 7. An investigation of the accidents of supersonic fighters shows the need for a twin engine configuration. Almost all the next generation combat aircraft are twins. This is no coincidence. Since fighter engines are "state of the art" ( sales jargon for "doubtful reliability") it is important to have twin engines so that the plane can get back to base. The number of Starfighters which crashed due to engine related problems is shocking to any investigator. The reported engine related failures in the MiG 21 is noteworthy.

Northrop once published a study where it was claimed that the peacetime loss rate of a twin engine fighter was one fifth of a single engine fighter. This would be true only for a very high and rigid standard of flying discipline but no doubt many of the losses are due to pure engine trouble. In the case of the MIG 21 crashes some 40-50 losses would have been avoided had the aircraft been twin engine.

8. The Chinese FC 1/FC 17 is probably the most intelligent development of the MiG 21 The lateral intake allows for a decent size of radar without obstructing forward view - something that declined with each new mark of the MiG -21. The high aspect ratio larger wing of the FC 17/ FC1would improve both low speed and combat handling as t would bleed off less energy and the lower wing loading would mean that the blistering landing speed would have been tamed to a reasonable figure. We should have had an alternative LCA project along similar lines. It is cheaper, more economical and faster to have two competing projects until a clear winner emerged.

Given the expected delay in the LCA induction the IAF could do a serious contingency study about rebuilding their time expired fleet of the MiG 21Ms. Aircraft are not like the fabled "one Horse shay" in that they don't fatigue all over at once. Life expired means usually the wings have no life left in them. Usually some 8 to 10 other components of the air frame suffer the rest are usually quite good for another 3000 hours. This would essentially mean remanufacturing the centre fuselage carry over structure and the wings and some of the empennage fittings. The RAF is "re-winging" its Hawks which is a good precedent.


1) The MiG 21 is a sound and excellently engineered design by one of the most respected design bureaus in the domain of fighter design.

2) The loss rate of the MiG 21 is in no way worse than any similar fighter of its genre and better than most.

Applying western accident rates is also somewhat unrealistic because of significant decay in thrust and lift due to high air temperatures. A 12% decrease in lift or thrust can lead to a 100% difference between crashing or getting back safe. There being no easy mathematical co-relation.

3) The design of fighters for mach 2 flight makes them difficult to handle during take off and landing. This is traditionally the most accident prone regime of flight- even in civil airliners operated by highly seasoned crews under vary benign flying conditions.

4) Statistically at least there is no conclusive evidence that poor training was a major contributor. Highly seasoned aircrews were involved in a significant number of accidents. Fighter flying is a hazardous job.

5) The ejection seat's performance left a lot to be desired. It may well have been worthwhile to develop a specifically tailored seat once the aircraft was switched to a role involving low level flights. Possibly the licensing agreements did not cover such a case.

6) The sudden decline in the crash rate of the MiG 21 après 2002 cannot be explained with the current level of published information.

7) Twin engine equipment and redundancy of systems, despite a higher first cost, may be essential in the future and may actually be economical during the life of the fleet

8) With innovative engineering the MIG 21 M and the Marut can be the basis of future equipment as a low cost supplement.

-Prodyut Das M.Tech M.Ae.S.I


The MiG philosophy of simple technology and advanced concepts

1. Some schools think of the Fighter as a showcase of technology, others differ.

The MiG Design Bureau, which set the world standard in high performance fighter design, for fifty years belonged to the second. The remarkable fact of the MiG fighters was the stead fast use of only proven ,decade old technology to achieve phenomenal performance which often the rivals equaled only by much more costly unproven and sometimes ineffective technologies.

The MIG 1/ MiG 3 used steel tube and wood structure in 1941. Despite a heavy engine it was the fastest fighter in the world with an unrivalled high altitude performance.

The MiG 15 used a ten year old centrifugal flow engine design copied from the Nene and yet it not only completely outclassed all other opponents ( Meteor, Panther etc ) having similar engines but also was a very worthy foe to the much more advanced F 86 Sabre.

The MiG 25 Foxbat used steel instead of Titanium and a simple engine that could be described as a large version of the Bristol Viper. Yet by clever but simple design of its intake system it was the fastest and highest flying fighter /interceptor/ PR aircraft for a very long time.

The MiG 29 was clearly set the standard in fighter maneuverability when it appeared yet it neither used FBW nor any composites initially.

A comparison of the design approaches of the MiG 21 and the F104

2. Of the five Mach 2 jet fighters - Draken, Lightning, Mirage III, MiG 21and the Starfighter the MiG 21 was the lightest, simplest and the most widely used. It saw combat in the Vietnam, Indo-Pak and Arab Israeli conflicts where it acquitted itself very well against much more sophisticated adversaries. A technical comparison between the MiG 21 with its adversary the Starfighter F104 A is interesting as it shows how superior packaging concepts permitted the avoidance of expensive technology.

In the F104 Starfighter was definitely the more sophisticated Aircraft. The brilliant Clarence Johnson, possibly spoilt by an abundance of technology and resources, used a combination of a 3.5% thick straight wing with a state of the art J 79c engine wit a 17 stage compressor giving a pressure ratio of 12. The 3.5 % thick wing required CNC milling which was cutting edge technology in 1950s. The wing was cutting edge in a literal sense also. On the ground the leading edge had to be capped to prevent damage and injury. Interestingly the intake was fixed geometry which meant that the intake was inefficient at off design conditions. The tyres, which could not be fitted into the thin wings were of extra high pressure and had to be fitted into a narrow track fuselage mounted undercarriage. The ejection seat was of a downward ejection type which must have been unnerving. The fact that the wings were almost solid meant that all the fuel was carried in the fuselage. This must have lengthened the fuselage considerably increasing its pitch inertia.

The MiG 21 team chose the innovative tailed delta concept. Initially derided by the West it was proved to be the best solution for the supersonic combat role. It combined lowest wave drag and yet avoided the problem of high induced drag of the pure delta which had to use "up" elevon" resulting in loss of lift during take off or a turn. Engineering wise the Delta plan form of the tailed delta meant a reasonably thick wing which could be manufactured by traditional sheet and rivet methods.

The engine compressor had only six stages but two spool technology (first used in the Daimler Benz ZKL in 1944 !) allowed a pressure ratio of 9. This combined with a conceptually sophisticated but engineering-wise simple translating intake cone allowed better ram pressure recovery. With typical MiG Bureau simplicity the cone was three position rather than being continuously variable. The overall pressure ratio was thus pretty close to the F104s but the fewer engine stages meant a much cheaper and lighter engine.

The MiG 21 is till date the lowest powered Mach 2 interceptor in service despite a profusion of bulges and scoops and having mushroom head rivets towards the rear.

The undercarriage was a brilliant design which allowed a wide track undercarriage with low pressure tyres for ease of ground handling. The ejection seat not only equaled the performance of the contemporary seat but the semi encapsulating feature gave an outstanding protection for high speed bailout Unlike many of its contemporaries all systems were duplicated.

The MiG 21 went on to successful service with both large and small, relatively obscure and new air forces which speaks well of its serviceability and reliability.

Whenever used in combat (Indo-Pak, Vietnam and Yom Kippur) the MiG 21was a very respected opponents to warplanes several times more expensive. Pakistan lost 3 Starfighters to the MiG 21 in 71 and the Israeli Air Forces had greater respect for the MiG 21 in the Yom Kippur war. The F4s and Mirage 3s usually avoided dog fights with the MiG-21.

Western Industry found the selling price of a MiG 21 unbelievably low and politically motivated but sheer good engineering and ruthless standardization kept prices down. Low prices led to mass production. If prices were indeed subsidized the amount would not be as much as is made out to be.

References and acknowledgment;

This is to thank Jagan Pillarisetti of Bharat Rakshak and Warbirds of India for allowing me to use the data on his Warbirds site as well as supplementing the data with further information and comments.

Wg. Cdr. Sekaran ( Retd) of MOFTU for his inputs in discussing the accidents listed in the Warbirds of India site.

The reader is also directed to search the following websites through any search engine. I used Google.

  1. Bharat Rakshak
  2. Warbirds of India
  3. Greg Goebels website on Fighter aircraft
  4. www.-916 - Starfighter for Starfighter losses and operational history
  5. Encyclopedia of Fighters Gunston for basic details of the Aircraft.
  6. Air International
  7. Air Forces Monthly
  8. mentions that a total of 315 MiG-21s were lost in about forty years - Oct 63 to end of July 2003 The MOD also stated that between April 1992 to March 2002 , a total of 102 MiG-21s were lost in accidents and 39 Pilots killed. We have records of 81 of these Mig-21 mishaps.

The opinions expressed in this piece are personal and do not reflect the opinion or policies of the organization I am currently employed in.

Copyright © PRODYUT DAS . All rights reserved