Talk:Crash of Malaysia Airlines flight MH17/Almaz-Antay investigation

In May 2015 Novaya Gazeta published online a confidential report prepared by Russian military engineers. Later the same material was published by Almaz-Antay.

Novaya Gazeta
Novaya Gazeta has published online a confidential report prepared by Russian military engineers. This is the most thorough forensic analysis of of the airframe damage so far. For a description of the damage, see our list of MH17 airframe parts. -- Petri Krohn (talk) 01:04, 7 May 2015 (UTC)


 * Sources
 * Это был «Бук-М1» (Report of Expert Engineers) - Novaya Gazeta, May 5, 2015
 * Rusvesna 06.05.2015 (summary)
 * "It was Buk-M1"--Classified Russian report on MH17 published - Fort Russ, May 9, 2015
 * Russian engineers’ report on MH17 crash ‘based on fake photos from Russian General Staff’ - UNIAN, 06.05.2015
 * ...please add more here. -- Petri Krohn (talk) 01:04, 7 May 2015 (UTC)

"Novaya Gazeta" (full) confidential document publication "The result of peer review of the investigation into the plane crash Boeing-777 (flight MH17) 17.07.2014 in the south-east of Ukraine", said the editorial. "To our knowledge, this work of Russian engineers MIC (including that of NGOs, which designs and manufactures missiles such as" Buck ") should be directed to the Dutch experts leading the investigation terrible tragedy", - said the material edition. The editorial said that felt it necessary to disclose the document in view of its social significance.

Conclusions (referring to "Novaia Gazeta" text)
 * 9М38М1 missile, 9Н314М warhead. There are 3 types of elements, heavy and light 1,2; heavy elements were I (double T) shaped ; light 1,2 are parallelepiped. For those dumbheads on internet: Figure 3A: wrong type, 3 B: correct type.
 * Figure 12 give their version of elements motion. 70 degrees horizontally, intersecting flight path from the right and back, 20 degrees vertically down. This indeed explains heavy damage to the left side of the cabin and give streaking damage on the wing and engine. So this is a good version matching many things overall. They do not address detailed shape of crater on the left wing, which appears to be more consistent with motion forward, not backwards.
 * Snezhnoe launch is inconsistent with the damage, illustrated in Figure 16
 * From angles of approach, they identified the launch area as Zaroshenskoe (to the south from the impact).
 * And by the way, this is where Ukrainian Buk was located.
 * Interesting Figure 9 (have not seen it before), with a comment: "Analysis of the damage elements of mechanization (flap) of the left wing can confidently say that it was mainly the field covering fragments. This is due to the presence of multiple lesions (striking elements of different fractions, light and heavy), and the amount of damage (density of holes per 1 sq. Meter, taking into account the distance from the point of detonation of a missile, and the effective area of the element is even higher than on the port (=left) side of the hull near the cockpit ).


 * They did not seem to really explain forensic-wise why, if coming at such angle, it was approaching from the South and not from the North. Seems to be it could be coming along that or nearby line  but from the North, not from the South (with direction modified a bit as it now moves closer to the plane, not overshooting it). This crudely gives their  location but reflected about the flight path, giving launch somewhere in the area of  Fasheevka. So the argument for the South seems to be really resting on the known Ukrainian Buk location.
 * Agree that Snezhnoe does not fit.
 * With approach from the front, missile moves relative to the plane approaching at 4 Mach; that gives some forward bias to the cone of elements; exactly how much bias depends on elements velocity. They do not discuss it though. On the figure, very slight bias seems to be shown (that would be true for element velocity much higher than 3-4 Mach... but it could well be also about 4 Mach, and not much higher...a bit strange that there seems to be no discussion of it). Elements flying off perpendicular at 4 Mach from the frontally incoming missile will move at 45 degrees relative to the plane frame...
 * --Resup (talk) 12:47, 6 May 2015 (UTC)

For any BUK theory to be valid, the direction of all projectiles should be traceable to a single point source. This would need to be somewhere near the upper left side of the cockpit. The report explains the narrow band of heavy damage by the shallow angle of the belt of shrapnel flying out of the warhead.

I am not convinced at the conclusions about the direction of the missile. It seems that the report assumes that the shrapnel flies sideways from the warhead. This does not take into account the Mach 3 speed of the missile. Combining the explosion pattern with the speed produces a forward directed cone. This 3D simulation describes such a cone and concludes that the missile came from 20 degrees to the right, i.e. from the direction of Snezhnoe. This direction is rejected by the Russian report. All this analysis assumes that neither the plane nor the missile turned before impact. If the missile came from ahead, then the wing damage could be explained by the missile body itself, not shrapnel. The left wind flap with apparent shrapnel damage may be a crucial piece of evidence. -- Petri Krohn (talk) 02:18, 7 May 2015 (UTC)

P.S. - The holes in the left wing flap are not BUK shrapnel damage. They have several orders of magnitude less energy than the holes on the side of the cockpit. (If these were BUK, then the cockpit holes would need to be armor-piercing autocannon.) More likely the holes are caused by debris from somewhere, maybe the missile. -- Petri Krohn (talk) 03:44, 7 May 2015 (UTC)
 * Most of shrapnel will stay behind blast wave, which slows down with the distance (e.g. there is a slow motion experiment here on cnn-very little gets in front of it, if any at all). Also shrapnel is incoming at ricocheting angles. Either of those makes low penetration not too surprising. But more than one event not excluded also. --Resup (talk) 13:36, 7 May 2015 (UTC)


 * Regarding video, cone of course will depend on both missile speed and projectile elements speed. Speed of elements is not in open sources; whoever produced the video had no clue and just draw some nice picture fitting the damage. This is not a simulation, as key info is not known and not easily simulated from what is known. That's OK for a youtuber, but engineers doing the report really should have addressed this at least. But with frontal approach, there will be considerably heavier damage in front, this is essentially what they said, and damage-amount-wise I agree.
 * Indeed, wing and flap damage are important overall --Resup (talk) 03:19, 7 May 2015 (UTC)

Direction Cones
If what they use has energetics of TNT/gunpowder, element speed will be under 2500 m/s (gunpowder), 3000 m/s (TNT)  by conservation of energy, and less than that in reality as some is going to get wasted. (Novaia gazeta mentions in one place a velocity of 2400 m/s for elements passing through towards the other side of cabin --this sort of matches up)  --Resup (talk) 04:09, 7 May 2015 (UTC)
 * Cone is said to be +/- 20 degree with respect to perpendicular to the hull direction, in the missile reference frame. Taking projectile speed at 2200 m/s, and knowing plane and missile speeds, allows to redraw the cone as seen in the frame moving with the plane. With Snezhnoe  (-20 degrees counterclockwise from the path), the cone misses the wing. With Zaroshenskoe (-110 degrees) everything works direction-wise.  20 degrees approach, kind of works, but impact is nearly perpendicular into the cabin. At 110 degrees, may be possible -along the wing towards the cabin, and half of the cone having a miss. Zaroshenskoe seems to be the best overall direction-wise fit; and Snezhnoe not working.  --Resup (talk) 07:29, 7 May 2015 (UTC)
 * I have tried to find sources for fragment velocity from HE shells and warheads. Little luck. Wikipedia does not even have an article for high-explosive shell. It would be possible to estimate the speed of shrapnel, if we knew the total amount of explosive, the weight of the shrapnel and the thermal efficiency of the design. In fact there is an online Gurney equations calculator on the UN site that does just this, using only the first two of the parameters, the "charge-to-metal ratio".
 * Gurney Equations for Fragment Velocity - UNODA
 * The results produced by the calculator do not depend on the size of the warhead, only the proportion of "Mass of Case" to "Mass of Explosives" and the explosive material used. With a 1/1 ratio of case to explosives the velocity is 2986 m/s for TNT and 3584 m/s and 3640 m/s respectively for RDX and HMX. The problem with the equations is that they give unlimited velocities at high C/M ratios, surpassing the detonation velocity of the explosive, although C/M ratios of under 2 seem to be safe. -- Petri Krohn (talk) 02:39, 9 May 2015 (UTC)
 * My numbers shall be called Gurney constant than--assuming that shrapnel is dragged by the blast and so explosive have to set itself to that speed --Resup (talk) 03:12, 9 May 2015 (UTC)
 * In wikipedia version--which references a book-- that safe m/c =1/2, cylindrical case, just means that v = Gourney constant; and according to wikipedia (probably = book) Gourney constant is 2440 m/s for TNT and does not seem to go over 3000 m/s for anything... UN calculator is messed up, (m/c) =0 is supposed to return 1.4 Gourney, not infinity; and infinite m/c supposed to return zero, but it does not. It should compute 1/( m/c+1/2 )^(1/2), but instead it computes (I think) 1/(m/c)^(1/2) + (1/2)^(1/2) . (Somebody please tell them it is not the same! Or somebody may break the speed of light limit!).  --Resup (talk) 16:21, 9 May 2015 (UTC)
 * Also broken: the number of opening and closing brackets must be equal.--Resup (talk) 17:36, 9 May 2015 (UTC)

The missile has a forward single point detonator. This causes the explosion front to travel backwards and imparts a rearward velocity component to the warhead fragments. The Almaz Antey presentation is confusing, but essentially the blast pattern is as shown in the diagrams for a missile travelling at usual velocity That is there is no forward vector component caused by the missile velocity. In actual fact most of the particles are thrown backwards relative to the missile vector.

This 'backwards throw' is not a significant penalty as slower particles are more likely to engage with the aircraft structure and impart energy compared to hypervelocity particles simply punching through without depositing much energy.

The Almaz Antey presentation makes perfect sense if you realise what they meant but perhaps got lost in translation.--Charles Wood (talk) 01:30, 15 September 2015 (UTC)
 * Detonation velocity is some 4000-7000 m/s so delay in detonation travelling front to back is negligible. Slight backward bias is possible; but in missile frame it is likely quite symmetric (as it detonates quicker than flies off). Cone computation assumed it is symmetric, no back bias.  --Resup (talk) 09:56, 16 September 2015 (UTC)
 * The striking fragments have a velocity 2000 to 2500 m/s. This is approx half the VOD of the main charge so the direction of the wave front will have had a significant effect on the direction of the particles. If the wave front velocity was infinite the particles would eject perfectly perpendicular to the charge. If the VOD was zero the ejecta would fly perfectly rearwards. However it's not infinite and actually similar (double?) to the ejection velocity, so there is a direction component of the ejecta that points significantly rearwards --Charles Wood (talk) 20:43, 17 September 2015 (UTC)

Other NG articles
18.05.15 Another publication, different author, of opinion that missile launched from NEE. --Resup (talk) 16:24, 20 May 2015 (UTC)

Novaia Gazeta published its third version, journalist investigation, now back to its opposition roots and blaming Russia (they want to have funding from all sources in the world or what?) No real new evidence is presented, replay of old things with some twists (old missiles left over by Ukraine, and Russian-origin launcher)...evidence is of sort that two unnamed persons made such a claim... but than one commentator is saying that no missiles were left over (in Luhansk, I guess)--so than this new twist appears dead on arrival or close to this... Saving the link; picked up by dw in an instant... --Resup (talk) 08:39, 15 July 2015 (UTC)

Almaz-Antay press conference
Russian concern Almaz-Antay held a press conference, aimed at quashing sanctions against them, to say that (if it was Buk, then) it was Buk missile 9M38M or 9M38M1 with missile warhead 9М314 or 9М314М1, because only those have X shape elements. That was taken off production in 1999, while the concern was formed in 2002, so they did not provide this missile to anybody; but Ukraine has it. Their conclusion is based on publicly available photographs; they also participate in the investigation and has exclusive competences. They say that Snezhnoe is inconsistent as that will take cabin off at once; offer to do an experiment to prove/establish who is responsible (an experiment seems right thing to do-and interesting how you account for movement)- interfax. They seem to repeat the version given in Novaia Gazeta (discussion), placing launch in Zaroshenskoe. They essentially say that some 40% of elements go perpendicular to the missile and so it needs to be on intersecting course to (have a chance to) account for all the damage. --Resup (talk) 13:33, 2 June 2015 (UTC)
 * Severe damage to the flap, illustrated in NG, with less damage to the wing, raises a possibility of a multiple event. --Resup (talk) 16:33, 2 June 2015 (UTC)
 * Their presentation is on youtube now in Russian and with English translation. --CE (talk) 15:11, 3 June 2015 (UTC)

Full presentation by Almaz-Antey ( Buk-m2 manufacturer ) about the incident. Still only in Russian. Unfortunately even in Russian somebody the diagrams are not self-explanatory. But main arguments are clear. It includes calculations of the attack angle of the rocket in order to cause such damage. - --Antidyatel (talk) 09:41, 9 June 2015 (UTC)

They say it was incoming at about 78 degrees; they reject Snezhnoe at 20 degrees; schematically in 2D cones are smth like this, with one extra alternative at -110 degrees shown --Resup (talk) 05:12, 9 June 2015 (UTC)

Speed calculation?
Resup, thanks for shifting my comment to more relevant location. I would like to add here one more analysis. Minimum Distance between Hit point (based on Dutch castrated report) and Snizhne is 24 km (Bellingcat magical tricks placed the launch spot even further). Buk-M1 missile goes at 860m/s, but let's assume 1km/s. Let's also assume that missile immediately travels at highest speed. For such distance the trajectory will be parabolic (slide 49 from Almaz-Antey). Than the time of flight will be way more than 30 sec and it will even intersect the pilot's vertical field of view. Dutch officially declared that there is nothing interesting on black box voice recorder. How is it possible for the pilot not to notice a 5 meters lobg rocket, that ejects violent plume behind, flies directly at him for more than 30 sec and even goes across his vertical field of view? And according to Dutch report, pilot was informed of another airliner in proximity, that he had to watch out for. So he was alert and paying attention to the surroundings. It is just impossible for the pilot not to comment on it, at least WTF. Even cloudy day doesn’t explain the issue. --Antidyatel (talk) 09:41, 9 June 2015 (UTC)
 * Last radio transmission at 13:19:56-13:19:59 UTC – MH17: ROMEO NOVEMBER DELTA, MALASYAN 17 p15, p29 (MH 17 speaking; RND=waypoint)
 * Both flight recordings ended at 13:20:03 UTC-- from the preliminary Dutch report (preliminary report discussion link)--Resup (talk) 16:22, 9 June 2015 (UTC)


 * The BUK missile burns out at around 14-15 seconds (from memory - it may vary but not much, it also depends on model, but I'm talking about the one alleged in this instance). The missile spends much of its flight in an unpowered ballistic trajectory but uses fins to make minor adjustments to strike the target.
 * The missile is much smaller than an ordinary aircraft so very hard to see. There have been numerous head-on collisions of large aircraft where the pilots never saw anything despite being on lookout. It's not surprising at all the MH17 pilots didn't see anything. --Charles Wood (talk) 10:11, 11 June 2015 (UTC)
 * We are in confused state with this -discussed at the end here. On some web pages, we can read about 15 seconds engine run, but there is a video out there which looks like old (long fins) model fired, and looks like firing engine for 40 seconds. Antey shows some trajectories which gives some clue, but not saying explicitly. --Resup (talk) 13:44, 11 June 2015 (UTC)

The 15 second for missile is from Wikipedia that quotes one Russian BUK that is about different military technologies. I've never seen it in some official document. Every source that talks about it eventually boils down to this book. But I strongly suspect that there is a typo.it should be 50 sec. Otherwise it doesn’t make sense. Missile is using proportional navigation and is supposed to chase manuvouting targets with speeds up to 830 m/s, while maximum missile speed of this kind is 860m/s. And it is supposed to catch those targets at distances up to 30 km. It just doesn’t make sense that after 15 seconds it is basically a stone with minor control indirection. Simple wind or requirement for slightly sharp turn will make it useless. Can you give an example of head on collision of planes? All I can find were either in the thick clouds or when one of the planes was ascending and another one during descending. They never crossed each other's field of view. Your comment about 5 meter rocket obviously relies on the assumption of 15 sec, meaning no plume behind. Youtube videos of Buk launches show plumes plumes staying way longer than 15 seconds. --Antidyatel (talk) 13:41, 11 June 2015 (UTC)

For those who still think that pilot of Boeing 777 is looking at the world through some kind of pinhole --Antidyatel (talk) 00:14, 12 June 2015 (UTC)

It is rare for pilots to see oncoming aircraft. Usually the first thing they get is an ATC warning or a TACAS warning. Gol Transportes Aéreos Flight 1907 is a classic example of planes heading directly towards each other and colliding without either pilot noticing anything --Charles Wood (talk) 08:17, 13 June 2015 (UTC)

Charles, it is an interesting example. Thank you. However, there was no crossing of the field of view, and pilots were not particularly looking around for danger, as testified by journalist Joe Sharkey on that flight "And it had been a nice ride. Minutes before we were hit, I had wandered up to the cockpit to chat with the pilots, who said the plane was flying beautifully. I saw the readout that showed our altitude: 37,000 feet. I returned to my seat. Minutes later came the strike (it sheared off part of the plane's tail, too, we later learned". It was 5 pm in Brazil, so it was already getting dark, but still bright enough to minimize contrast for light indicators on incoming jet. I would imaging pilots were relying on electronics, to tell them that something is wrong. In case of MH17 Pilots were informed by ATC of incoming traffic. So they were alert. Looking at this problem I also found out an interesting thing. When we say that plane was at particular hight it is no based on direct hight measurement but on response of barometric sensor at this location. Although the radar will give a direct measurement. So there can be conflict between what ATC operator instructs tot he pilot and how pilot controls the height. . Not sure how do they deal with such issue. What if the pressure sensor is faulty?

P.S.: I hope you are convinced that 15 seconds for BUK missile is not reasonable. --Antidyatel (talk) 15:49, 13 June 2015 (UTC)
 * Manual gives 19 to 23 seconds engine run... (links)--Resup (talk) 08:24, 15 July 2015 (UTC)

September, 2016
Presumably another release by Almaz Antey. A brief entry for now. Note page 9: only Buk M1, 9M38M1 missile, 9H314M1 warhead, have bow-tie fragments. Those are 13x13x8.2 mm, and 8.1 g weight (after impact may be a bit lower but not a lot). 1870 of those in the warhead, and there are two other types of pre-formed fragments in that missile. Also noted that investigators have a messy story on bow ties, claiming some were found, but too light, and (apparently) changing the weight subsequently. Weighting photo on early find, page 24 of this Almaz report, shows 5.5 g weight; this is Gilze-Rijen, Netherlands, Feb. 2015. In the published investigation report, October 2015, p.92, two bow ties listed, cockpit, 6.1 g, and human remains (pilot), 5.7 g. Also listed is 'cubic', in human remains, 12x12x1, 1.2 g weight. That's strange as it looks as a slice of a big cubic fragment from another warhead, 9H314 (13x13x8 mm, 10.5 g), and such size cubics are not present on other warheads. On 9H314M1, there are 8x8x5 cubes, and turning those into 12x12x1 would be 'rather difficult' (and it's 0.45 of volume). There are also 6x6x8 posts on M1, need to hugely deform to make into 12x12x1, and it's 0.48 of the starting volume... --Resup (talk) 21:56, 14 September 2016 (UTC)


 * Публикуем текст, который может быть положен в основу российской реакции на выводы Нидерландов об обстоятельствах гибели рейса МН17 - Novaya Gazeta, September 14, 2016
 * ''We publish a text that can be the basis for Russia's reaction to the Netherlands conclusions about the circumstances of the death of MN17 Flight
 * Bellingcat ответил на доклад «Алмаз-Антея» об обстоятельствах гибели рейса МН17 - Novaya Gazeta, September 14, 2016
 * ''Bellingcat responded to the report of the "Almaz-Antey" on the circumstances of the death of MN17 Flight
 * English-language video highlighting some of the issues raised (many tangential hits on MH17, fragments mass, shape), 16 Sept. 2016. --Resup (talk) 16:00, 16 September 2016 (UTC)

2021
Echo Moskvy are dumb-ass shitheads. Lukashevich shows/rebroadcasts a good video of the court looking at the primary (small) and secondary (large) fragments on the right hand side. Yes, there are some, and it is some problem to an extremist interpretation of Almaz-Antey report. But fragment angles relative to missile is not simply 90 degrees (perpendicular to axes); Almaz-Antey gives it as [64 to 124 degrees (with zero degrees would be straight forward, and in blurry font, a bit hard to read). In the reference frame of the plane, missile is moving, as well as the fragments, and those velocities needs to be added as vectors, so it gets biased, somewhat, from those missile-framed numbers, with more stuff flying in the direction of missile flight, compared to stationary missile. (Spread of fragments velocities also needs to be accounted and broaden angular spread somewhat). With all this, fragments may hit the right side either way, but most of the fragments will hit if it something like Sneznoe, and few, from rear part of cone of directions , if it is something like Zaroshenskoe. Almaz did not see good pictures of the right hand side--this is shown for the first time I am aware of, and apparently mis-stated that none hit there. But this is not the ultimate decider; they did show  damage from other fragment flying along plane axes, and one needs to compare which way most fragments went, not just whether or not any damage is present. One of their points is that Dutch modeling has rear angles wrong by 15 degrees, and so best Dutch match needs to be rotated , like 15 degrees or so, and with this, Sneznoe is not the best match for the location.

Cones shown on this page give some idea (not accurate one, as I do not have all parameters and was using rough estimates). Don't remember all that well now; angles on figure must be clockwise; more stuff flows in the direction of missile motion, clearly. There are some gadgets on internet, somewhere, which allow to do similar plots interactively, inputting your own parameters. --Resup (talk) 00:37, 30 May 2021 (UTC)

Moreover, Almaz experiment (Snezhnoe hypothesis) show penetration in the upper- and-front part on the right hand side, close to and a bit behind the cabin windows. While prosecution looks at holes further back and down (26:13 --31:24) mostly. Which would be closer to "along plane axes" then "cutting across" (while in both cases there would be a spread of firing angles, and deflections after impact). This needs to be carefully considered, long ago, and not just by lawyers pointing their fingers and measuring tapes --Resup (talk) 07:58, 30 May 2021 (UTC)

Radar data "found"
In a press release on September 22, 2016, the company announced that they "found" relevant primary radar data from the day of the crash that was hidden in their archives due to quality control reasons. They handed it over to Russian authorities who will hand it over to the Dutch-led investigation. --CE (talk) 17:54, 25 September 2016 (UTC)
 * see Talk:Crash of Malaysia Airlines flight MH17/Radar data