The HPT 32 Crashes - An Alternate Logic

Professor Prodyut Kumar Das

The recent article in Vayu V/2012 Requiem for the HPT 32 was carefully researched and provoked thought.  The fuel divider and the Collector tank location and capacity are the prime suspects. However there appears to be some uncertainty; The article mentions that at one stage the Fuel Divider was taken off the list of suspects by the investigators and there was the case of an engine stoppage on the ground. There is a certain uncertainty.

The uncertainty could be from the following. It is true that unless the fuel supply is smartly cut off, a warm fuel injected engine will continue to “diesel” even after the ignition is shut off. This “cut off” is one of the major functions of the fuel divider. However this function “gets out of the way”, so to speak, when the throttle is opened beyond idle or near idle. If the fuel divider is the culprit all the seventy odd incidents of engine stoppage would have occurred at idle or near idle conditions. Also, if maintenance is a problem NO failures should have occurred with a new/ “not overhauled” fuel divider. Has this been indeed the case? 

Regarding the fuel pipe line being not as per FAR standards the usual requirement is that the pipeline should be able to handle one and a half times the TO fuel flow rate. For the engine in question the TO fuel flow is approximately 1.3 litres per minute or 20 ml per second corresponding to fuel flow velocity of about 0.7 metres /sec which is also not too bad. In any case the main restriction to flow would be the filter just upstream of this line and increasing the pipe diameter will not make a decisive difference. Mind you at the idle case the fuel flow would be around 200 ml per minute so both the usable header tank capacity and the pipeline would be unlikely to be a prime suspect. By my estimate, even with the usable 3.5 litre capacity quoted  the engine could chunter on for a quarter of an hour at flight idle or three minutes full chat –both times more than enough to get the aeroplane at least into level circuit if not on the ground. That too under flight conditions of no bank or turn during the descent because a bank would recharge the header! It is also to be noted that there was an engine stoppage on the ground-when the aeroplane was near as level and  feed / collector capacity problems could not have occurred. Finally it is bemusing to accept that a failure rate of 77 failures in 400,000 flight hours-that too in a system that is “on” every minute of the flight- if a single component or system is at significant fault. In my view there is clearly room for an alternate hypothesis.

Let me say my Mea Culpas right in the beginning. I had no chance to see the aeroplane or have access to the data and the hypothesis is based entirely on conjecture and my experience as an Engineer and I.C. Engine man. The starting point of the alternate construct is that considering its usage the rate of failure is very low. Could the failure be due to the fact that there are a fairly large number of random factors which almost never occur at the same time. When they occur together, however, they cause an engine “failure”. The “rare random combination” better explains the one failure every 5000 hrs. What could be these factors?

a)      The poor engine is suffering from the “Glider tug” syndrome. A high power –low air speed combination as in towing gliders have been known to cause engine failure. The HPT 32 does not tow gliders but it is an extremely “draggy” aeroplane.  If you visually compare the HPT 32 with the similarly powered SF 260 you will get the point. If you are one of those who will point out that the Italians will get “style” into concrete “tie down” blocks then look at the Finnish Vinka or even the Bravo or the Bulldog. In the HPT 32 the contours of the cowling and the canopy, the untidy undercarriage linkages and the huge fin hurt the eye. One must mention the oversize fin. The excess wetted area results in excessive parasite and induced drag leading, again, back to an overworked engine. The engine has to operate at a few notches higher throttle setting compared to other installations and yet not get enough cooling air.

b)      The cooling of the last row of cylinders in a horizontally opposed engines requires, as the Germans say, “Patience, experience and maturity”. Particularly the rearmost cylinder opposed to the direction of the propeller rotation is, cooling wise, in a severely unfriendly environment. Thus the fitment of the cooling baffles and its maintenance is of greater than usual relevance in this case.

c)      In India the cooling air itself is 20 to 25 degrees higher than ISA. This would rob it of about thirty percent of its cooling capability. Bidar is notably dry.If this is combined with the occasional less than “normal” humidity we can see problems lurking around the corner. I dare say that if the HPT 32 operated above 35° N we may not have seen this problem at all!

d)      The dust and the dirt. The metered fuel supply system takes input from the static and rams pressures. If this is not “klim bim” perfect then the mixture would lean out to the point when the engine would starve and stop. Dust would also reduce the cooling heat transfer.

So what could be happening? We have an older (somewhat dented and battered and the cowling and canopy rattles a bit in flight!?) aeroplane flying a sortie on a dry dusty day with some prolonged spirited flying at high power. The engine is hot. As the power is reduced and that aircraft is gliding back the cooling flow is reduced by the low airspeed; the heat accumulates under the cowling. May be the baffle seals are just a little aged. All add up to –in those rare occasions- leading to a local overheat, distortion and “incipient seizure” in the engine. The high oil temperature and hence reduced viscosity of the lube oil would be additional contributor in this construct.

For reasons too boring to detail here I once had a car that had 90,000 kms on the clock. It had this trick of the engine suddenly “seizing” yet when I let the old girl be for some time -this was in Daman where chilled liquid coolants for me were easily accessible!- it would restart as if there had been no quarrel. There was another case when a students’ designed racing car that would stop suddenly due to over heating. A better designed duct for the radiator cured the problem very satisfactorily. Perhaps our engine is having the same problems?

Incidentally the HPT 32’s wing and span loadings are some 40% higher than the HT2s and so the glide ratio and minimum airspeed would be that much poorer. At low altitudes the pilots have that much less a chance of a safer landing or a pancake.

If the above construct is a possible model then what is to be done?

In the immediate term:

1. To increase the routine maintenance quality. The baffling of the engine is a prime suspect and so must come for close inspection at suitable intervals. Cylinder fins and the static and pitot ports for the AF system to be checked for dirt and should be inspected for cleanliness as per flying conditions. I mention routine maintenance and preflight checks only because a recent issue of a NTSB bulletin mentions fatal crash at take off killing six people because the pilot had failed to drain the fuel tanks of accumulated water. In his previous company someone else used to do this for him.
2. The quality of fit of the cowling and the canopy joints and panels to be improved by the fitters to the extent possible to reduce drag. Older airframes to be examined for the usual dents, bumps, loose fittings with the above prognosis in mind.

 In the medium term the following studies to be made.

3. An OR study into the accidents based on the above assumption that “an unfortunate combination of circumstances” rather than major system fault is the cause of the “engine failure”.
4. Initiate the design of a neater cowling and canopy. The cowling lines of the SF 260- which incidentally has the same engine - is the work of a Past Master (Stelio Frati) and could be an inspiration. These could be retrofitted at the FTS .
5. Do a CFD study of the through flow and back flow on the oil cooler. I have seen significant improvements with some very simple “fixes” and better lubricant viscosity would be a definite palliative.
6. Do a study to find out how difficult it would be to fit a semi-retractable undercarriage as in the Yak -18 and if there would be any benefits.

The real “de luxe” solution is economically unviable but is mentioned for the completeness of the discussions. It is entirely a personal view that HAL spoilt itself by the success it had with the big fin to improve the spinning characteristics of the HJT 16 which I have seen has very reliable spinning characteristics. What worked for the Kiran was possibly tried again on the HPT 32 but the balance was lost. Optically the HPT 32’s fin is huge resulting in excessive weight and drag. Others rely extensively on strakes to generate flow across the fin and rudder in a spin. The German Grob is of course  typically Teutonic in its determined application of strakes and under fins but the Bulldog, and the Vinka not to mention the SF 260 all use strakes quite discreetly with success to achieve  desirable spinning characteristics. The gains of using strakes for good spin recovery are in weight and drag which seems to be the root problem here.

Prodyut Das

Professor

Prodyut Kumar Das is an Alumnus of St.Xaviers’ Hazaribagh, IIT Kharagpur, and IIM Kolkata. He started his career with Aircraft Design Bureau HAL and for twenty years worked and led various vehicle related Product Development Projects with leading Indian and multi National Companies.

He left Industry to join IIT Kanpur in 1993 as a Professor in the Department of Mechanical Engineering. There he won a prize of the Royal Aeronautical Society of UK for his design of a light sports aeroplane using grants given by ARDB. He also did a project study on “The design of a Light Car costing less than 1 Lakh” which was a Ministry of HRD funded project IDICM 36 and started his research on Stirling Engines in which the IN was keen.

When IIT Kanpur did not renew his 5 year tenure he returned to the  Industry as a Vice President Technical and finally retired as Advisor Aerospace in the e- Engineering Division of a Leading Indian Engineering Company.

He currently teaches Engineering in a Private Engineering College in his hometown and continues his Research as a Consultant. He has been writing on matters related to Defence Engineering since 1990s.

The Ails of the LCA

Professor Prodyut Kumar Das

Kolkata, June 2014

I wish I had a guinea every time ADA missed out on a Date. I would have been, if not rich, at least well –to-do. I say this because recently, this last December , I think, one of the key figures of the programme- one might say- the Father of the LCA- stating that we would have two LCAs more by March and, if I remember a right- half a dozen before the year is out. The Ides of March have come and gone, “April, the Cruellest of Months” has gone and now even “the Darling Buds of May” have wilted. “June is ready to bust out but nary a sign of them those Airplanes!

It is worrisome when ADA repeatedly fails on dates because these are symptoms of  cluelessness. The highly qualified gentleman in the above paragraph must have had access to the proverbial “Horses Mouth” and yet, not for the first time, he has been hopelessly wrong. Is it really so difficult to predict the future events?

 In India we have a culture of very accurate predictions based on informal methods and folklore. The apparently “stupid” farmer kicking the dust as he chews slowly on a dry rice stalk may predict the Weather quite accurately. The old Crone sitting under the Neem Tree as she berates her newest daughter- in -law will still be able to predict whose Bahu is going to be a Mother -sometimes even before the poor girl herself is aware! Neither the Farmer nor the Crone has any “scientific” qualifications but they still come pretty close to the truth. So why not try applying those techniques on the possible date for the LCA?

Let me say before I begin that I have no access to “inside” facts. I am a very seasoned Engineer and I like machinery. That is all. What I am writing is therefore a construct. Of course ADA may, (out of sheer spite!) come out with a squadron of LCAs by December along with a chorus of well trained mechanics and a well organized stream of spares etc just to prove me wrong. That would be quite nice. In any case if people speaking from the Augean stables are so repeatedly wrong I am in “eminent” company if my here predictions are wrong. Of, course, mark my word; I fear I shall be proved right!

Let me put down the more important tasks remaining for the LCA to get FOC.

a) Opening the full envelope of positive and negative ‘g’

b) weapons firing particularly of the 23 mm GSh.

c) Spin trials

d) Missile Launching.

e) Proving of remaining systems.

Missile Launching: Pakistan managed to jury rig the AIM9 onto the MiG19 in a matter of months there is no reason to expect that the same cannot be done onto the LCA. I am referring only to CCMs. It will be a brave Air Marshal who will refuse the LCA solely because the aeroplane cannot fire BVRs for the moment.

GSh 23 firing.

The problem of gun firing is “old hat”. The Chemistry is Class 9. The gun propellant gases are ingested by the engine and that affects the air fuel ratio as the propellants gases displace the oxygen in the air causing the engine to flame out through “over richness”. This is aggravated by the pulsations of firing which will tend to “blow out the candle”. This is particularly true at high altitudes where the air is “thin” causing both effects to be amplified. The old trick is to “dip” or reduce the fuel to the engine automatically when the gun is fired. In the LCA, a one second burst will release about five kilos of gun gases into a region of inlet flow of 4 kilos of air over the same period at high altitudes. Vibration is of course a problem but the GAST system (look up!) of the GSh 23 means the recoil loads are much less. I do not think the horrible memories of the HF24 -where I still believe the concerned German Engineer probably put a “bug” into the design- will be repeated here, especially if ADA has had the wit to use the forged aluminum cradle or its derivative the MiG Bureau used for the MiG 21M’s mounting.

24⁰ AOA

This is the old Phantom joke now gone sour. I would like to meet the person who will refuse the LCA simply because the aircraft won’t do 24^0.

Proving of remaining systems

Thirteen years after the first flight there would be very few things that require major tweaks so there is very little that remains to be done.

Does that mean then we can see a FOC by December and a steady stream of LCAs from 2015. No, definitely not, because I guess the Mk1 is still a ”lemon”. It is not combat worthy. I am on shaky grounds here because I am making the previous statement entirely on what is available in “open source”. The LCA was “officially” declared to be about 1300 kilos overweight by ADA. Subsequently there has not been any announcement about the weight being corrected. Certainly the weight correction would have been noised about. If you have “inside” confirmation that the basic empty weight of the LCA is around 5100 kilos don’t read the remaining portion because everything written below is then irrelevant.

Why is weight important?

Airframes will tolerate a fair amount of abuse but they cannot tolerate excess weight. Let us take the MiG 21 Bison. Despite its age it is still relevantly “sprightly” as Cope India showed. The MiG 21 is of the same thrust to weight class as the LCA. Now imagine we poured in 1300 kilos of lead (Plumbum!) into the airframe. Immediately all critical parameters- take off run, acceleration, climb rate, radius of turn, range, ceiling and top speed will fall below current designed figures. In short the MiG 21 will not be fit to fight. In summer thrust and lift reduces by about ten percent and things would be worse! Exactly the same is happening to the LCA. Until the weight has been corrected the aircraft cannot even complete its flight test programme. My Farmer’s guess is that ADA should have an airframe weight of around 2300 kilos and an undercarriage weight of around 300 kilos to come out shouting winners. Mention has been made of the LCA requiring ballast. Aeroplanes sometime require ballast to get the CG right. The HF 24 needed 134 kilos about 2 % percent of the basic empty weight. This was in the days of wooden slide rules but evidently someone cared. How much ballast does the LCA need? Given the use of CAD it should perhaps be no more than half that figure.

“Opening up the envelope”.

This cannot happen safely because the “g”s to be applied requires acceleration and lift. Unfortunately lift means drag particularly in AR Deltas whose induced drag is almost double of comparable swept wings. Given the combination of excess mass and drag the F 404 just may not have enough “urge” to pull the little aircraft around a turn at 8.5 G i.e. the aircraft is power limited and lift limited to pull the required “Gs”. One could of course dive the aircraft and do pull ups but I think it would be a pointless exercise because one would have to do it again when the definitive airframe is available.

Spin Trials.

This is also held up because of weight. A spin is a combination of a stall and a turn at low airspeeds. The aircraft sinks because of the stall and it yaws and rolls (slowly) because of differential lift and drag caused by the different airflows due to the turn over the two wing panels. The forces at play are the above aerodynamics loads and the inertia of the aircraft which depends on the weight of the aircraft. Given these basics the LCA will be reluctant to spin because the Delta wing is usually difficult to stall. Given the excess weight/ inertia it will take a long time to stabilize the spin. Height loss in recovery will be “interesting”. It may be recalled that the Mc Donnell Phantom II was so difficult in spin recovery that if the crew had not recovered from the spin by 10,000 feet the drill was to eject. Well that is a precedent anyway!

So unless you have tackled the weight you can’t do the spin trials. What happens to the FOC? Please do try and not have FOC 1,2 etc.

Intake Problems

There have been persistent reports of “intake matching” problems. What happens is the take off requirements of the intake are in direct contradiction to those required in transonic flight. You either accept poor take off and climb or face high spillage drag and engine surge at transonic speed. The solution is conceptually and mechanically very simple. Aeromodellers flying ducted fan models (PE Norman’s ducted fan MiG 15 of happy memory!) used them. We used to call them “cheat intakes”) .Spring loaded “blow in” and “dump out” doors are generally used. Even the dear old Hunter of Good Queen Victoria’s times (well, almost!) had them. You could see them on the wing intake lips. As I write about this I realize that I have not seen any photos of such doors yet on the LCA . Perhaps some reader can post?

Aerodynamics

I have elsewhere mentioned that the LCA is aerodynamically blunt, its comparable equivalents being almost a meter longer. Any Aerophile will remind you the Douglas A4M with the 10% more power was actually 0.1 Mach slower than the less well powered Hunter Mk6 which had a longer fuselage and better entry Supersonic wave drag depends on the maximum cross sectional area and its position along the longitudinal axis as well as the entry aerodynamics i.e. from the radome tip to somewhere behind the rear cockpit bulkhead. ADA needs to go over the contour and the cross section centimeter by centimeter. I am not exaggerating because it is so easy to end up with excess weight and wetted area if one becomes too enthusiastic. It is not for me to dare suggest but for God’s sake use some “feel” along with the Analysis.

Maximum speed.

My betters have said that the aircraft has reached Mach 1.4 -(or was it Mach 1.6?).). Sorry, Guv’nor but the facts don’t tie up! We seem to have on our hands an overweight aeroplane that is significantly stubby and has inlet problems and yet it reaches its design speeds? Cap in hand, with fingers touching my forelocks (Alas! Long gone to happy hunting grounds!), I would say no, Sors, this bain’t true! What may have happened is that the claimed speed has been achieved in a dive of around 30⁰.

The Prognosis

Common sense is that if the LCA Mk1 is reasonably well designed it should be in the same class as the early Gripens i.e. definitely superior as a replacement to the early Mig 21s which have begun to retire. The LCA Mk1 should be clear for super priority production. Somehow that is not happening and, going by precedent- not going to happen. The horrible suspicion is that we will see only “token” numbers of the LCA Mk1on v pretexts of manufacturing difficulties etc as a rearguard action until, hopefully, the LCA mk2- which will be an almost new airframe design, - is ready. We will, of course be relying on an organization, which could not correct an overweight problem it itself acknowledged in near twenty years (1996-2013). I am so glad I am not the Air Chief!

The interest expressed by the IAF in the AJTs is perhaps a corroboration of the above. The YAK 132 is a fairly useful LCA if you look at it carefully and indicates how little was actually wanted by the customer before ADA went gaga over Technology. Reminds one of Tacticus who had said so long ago “The enthusiasm for war is highest amongst those who have the least experience of it”. Replace “war” with “Technology” and you have the gist of the situation.

Prodyut Das

Professor.

Prodyut Kumar Das is an Alumnus of St.Xaviers’ Hazaribagh, IIT Kharagpur, and IIM Kolkata. He started his career with Aircraft Design Bureau HAL and for twenty years worked and led various vehicle related Product Development Projects with leading Indian and multi National Companies.

He left Industry to join IIT Kanpur in 1993 as a Professor in the Department of Mechanical Engineering. There he won a prize of the Royal Aeronautical Society of UK for his design of a light sports aeroplane using grants given by ARDB. He also did a project study on “The design of a Light Car costing less than 1 Lakh” which was a Ministry of HRD funded project IDICM 36 and started his research on Stirling Engines in which the IN was keen.

When IIT Kanpur did not renew his 5 year tenure he returned to the  Industry as a Vice President Technical and finally retired as Advisor Aerospace in the e- Engineering Division of a Leading Indian Engineering Company.

He currently teaches Engineering in a Private Engineering College in his hometown and continues his Research as a Consultant. He has been writing on matters related to Defence Engineering since 1990s.