Talk:Crash of Malaysia Airlines flight MH17/Forensics

Direction of projectiles?
In any Buk theory the shrapnel would have to come from one single direction or source. One attacking fighter would also produce "machine gun" fire from a single direction. Some people see both entry and exit holes in the same segment and thus claim that two different fighters attacked the plane. I believe that all the major holes can be explained by projectiles coming from one direction.

The direction of projectiles seems to be from the front, slightly to the left and slightly above, directed mainly at the captain's chair through the windshield. The projectiles left perfectly round holes of about 20 mm diameter in the upper frame of the port side windshield. Further down the left side the holes became elongated and irregular, some with the appearance of exit holes. The smaller irregular holes on the side are most likely the result of fragmentation rounds mixed in with armor piercing bullets. Some of these may have bounced off the side and exploded outside.

The same direction of the projectiles is also evident in this analysis of the hole in the cockpit roof / side pannel done by the Bellingcat team.


 * You reference the "Bellingcat team". Unless you can you can reference their source material with usable links and evidence of good provenance then such a reference is not useful in any way. Bellingcat is by self-description an agency of Eliot Higgins and uses a paywall to access material so no links are possible. Beyond that the reliability and objectivity of Eliot Higgins is very low. Can you provide original sources please.

The damage to the left wingtip was most likely produced by debris from the cockpit section hitting the wing (as Charles already suggested). -- Petri Krohn (talk) 04:04, 6 August 2014 (UTC) Updated 08:43, 7 August 2014 (UTC)

Update on wing damage: After looking at the KLA902 wing damage I am more inclined to believe it was caused by a R-60 or similar air-to-air missile. It is not breaking Occam's razor to assume that a jet would use both the autocannon and missiles to bring down MH17 – in fact it would need to do so, if it was to guarantee the success of its mission. Also it is possible that the apparent blast damage and random shrapnel holes on the front port side is caused by a missile exploding and not the autocannon. For arguments sake we can dismiss this. Supporters of a Buk hypothesis can always argue other causes for the wing damage. -- Petri Krohn (talk) 08:20, 7 August 2014 (UTC)


 * The wing, okay - it happened while airborne, but it could well be that. I previously thought the left-side damage looked to have a from-behind trajectory. Looking again, it looked more straight in. Looking closer yet, as I show below, I think these holes come from forward, a bit above, and at a pretty oblique angle - obviously a bit to the left, but not much. In line with what you just said, Petri. (from left side, forward cockpit, just below the windscreen - detail from the high-res photo available here) --Caustic Logic (talk) 10:28, 6 August 2014 (UTC)


 * I was also just thinking how that ubiquitous graphic lining up the wing damage and the cockpit throws everyone off. We can whip up an antidote, right? Which images should you show me in review and let the reader follow along? --Caustic Logic (talk) 10:28, 6 August 2014 (UTC)

Effects of R-60 air-to-air missile on Boeing airliner?
Korean Air flight 902 is extremely relevant to the MH17 case: The plane was forced to land by damage from one R-60 missile There are striking similarities between the damage and those seen on MH17: -- Petri Krohn (talk) 17:57, 6 August 2014 (UTC)
 * The Korean Boeing – English Russia, June 3, 2008 (photos)
 * R-60 Missile Damage – DeepResource, 06 AUG 2014
 * Korean Air Lines Flight 902 – Wikipedia
 * Photos: 1, 2
 * 1) Four meters of left wingtip cut off
 * 2) A pattern of shrapnel damage on the side of the fuselage where the warhead exploded
 * 3) Large sections of the fuselage torn open

Air-to-air missile or 30 mm autocannon?
The damage will tell us it was an air-to-air missile with rod warhead fired from the front, that exploded when its proximity fuse passed the cockpit window. -- Petri Krohn (talk) 23:04, 28 July 2014 (UTC)


 * More specifically,I don't think the angle of impact can tell us anything. Maybe I was the only one that for a minute thought it could - if it scraped the top of the wing, it must've been fired from above the wing, behind and to the left. But it could have just leveled out there after flying up, turning, etc. Other clues are what they are - payload type and size, dynamics I don't know. I note Charles here said the damage looks less like a rod than regular shrapnel, if that matters. For my part, I'm suspecting Surface fired, partly because the people who should know (Kiev) keep saying that and claiming they can show you the exact launcher. Maybe they can and did. Maybe the jet was there to help explain why the separatists did it, and to observe, or something. --Caustic Logic (talk) 23:14, 29 July 2014 (UTC)
 * This analysis say it was a 30 mm autocannon. Maybe. (English summary + another) One thing is sure; it was not a Buk or fired from the surface. -- Petri Krohn (talk) 23:22, 29 July 2014 (UTC)
 * Russian analysis says it is not a Buk. -- Petri Krohn (talk) 23:36, 29 July 2014 (UTC)


 * Since that's what I suspect at the moment, some discussion should follow. Will check those links soon. --Caustic Logic (talk) 23:42, 29 July 2014 (UTC)


 * Effect of 25mm and 30mm cannon shells Gives an extremely good idea of what cannon shell damage looks like. I don't see any damage pattern on MH17 wreckage to match that. On the other hand the damage pattern from a proximity detonated shrapnel warhead is entirely consistent with the damage. E.g.


 * [[File:Shrapnel_sample.png]]


 * An expanding rod warhead is eliminated as there is no damage match. E.g.


 * [[File:Hobartdamage2.jpg]]


 * Damage to the left wingtip is variously being touted as an R60 air-to-air missile expanding rod strike or a 30mm cannon strike. It's not an expanding rod strike as it scrapes rather than penetrates the skin in a line. It also does not match a 30mm cannon strike as there is no detonation pattern and 30mm rounds don't 'bounce' as they skip along a wing surface. It's suggested that this damage is actually from later debris falling on the wing.


 * [[File:Wingtip_damage.png]]


 * --Charles Wood (talk) 03:39, 30 July 2014 (UTC)

Left Wing Damage
(copied from above to start this section) Damage to the left wingtip is variously being touted as an R60 air-to-air missile expanding rod strike or a 30mm cannon strike. It's not an expanding rod strike as it scrapes rather than penetrates the skin in a line. It also does not match a 30mm cannon strike as there is no detonation pattern and 30mm rounds don't 'bounce' as they skip along a wing surface. It's suggested that this damage is actually from later debris falling on the wing.


 * [[File:Wingtip_damage.png]]

--Charles Wood (talk) 03:39, 30 July 2014 (UTC)


 * Wing damage after/during crash: I thought that too on first look, but based on nothing. Since then I decided the irregularity of the windward edge of that tear suggests it was rippling with airflow as it was torn, so it was airborne at the moment, anyway. That's a lot of scattered, forceful, linear action lining up with the cockpit to be coincidental (brush damage?) - I think this is munitions damage. Another gouge I found interesting is a more likely mechanical cut from the plane falling apart after the attack. Seen closer up and seen in context (just southeast of picture's center) --Caustic Logic (talk) 09:58, 30 July 2014 (UTC)


 * Wing damage appears consistent with projectiles moving towards the plane, not away from the plane.
 * Depressions are towards the wing back edge, glancing scrapes are towards the front. If motion is away from plane, there is no escape root from the depressed part. Thus I feel hit is coming from the back-left, moving towards the pilot cabin. This appear to contradict an assumption of front-to-back hit (but not the overall conclusion)  made by a Buk expert.
 * Agreed that some of the damage towards the wing front appear to be due to incoming air drag. Such drag, subsequent chaotic fall, may explain why wreckage is not aligned with the plane course.
 * Very unlikely that damage is not due to a projectile of some sort at all: a lot of relative speed and glancing blow. --Chingachgook (talk) 22:21, 9 August 2014 (UTC)


 * Hey, I think I get what you're saying, agreeing with what I said. But since then, it's come to appear the blast/attack (if only one) was from the front and this wing damage is coincidental. I still thenk it was while airborne, but there was some breakup then. Direction, I'm not sure. It should be away from the plane, if it's plane fragments that scraped across here. It seems like the forward part is what came apart early on, and any piece of it could cut the wing open as it nose-dived past the chunk with less mass and more drag. So, I can/have seen it both ways. --Caustic Logic (talk) 23:42, 9 August 2014 (UTC)


 * It was, kind of, similar to stone skipping. I expect smoother, more shallow part of the dent to point in the direction of the motion, that is towards the plane as detailed dent photos suggest. Should not be difficult to do experiment, sand or something, to confirm. This also suggests nearly horizontal impact angle. (All of that does not look consistent with front or right- fired Buk) --Chingachgook (talk) 05:14, 10 August 2014 (UTC)

Not a Buk?
Responding to Petri's non-BUK analysis links: the German analysis - the second guy sounds fishy. Most of the damage looks too irregular to be bullet holes. But then some of them, especially on the left here, do look consistent to me. I'm no weapons expert. It's possible it was hit with two different things at different times. (image below is huge - full size in new window gives some great detail) --Caustic Logic (talk) 09:49, 30 July 2014 (UTC)

Okay, the Russian analysis - Google translated: "The probability that a passenger plane was shot down by a missile fired from "Buka", is negligible, said Maj. Gen. Cruz in an interview with "Military-Industrial Courier."
 * "All openings of submunitions on the paneling" Boeing "suggests that undermining warhead was done bottom-side. But our "Bukovskaya" missile hit the top, making the "hill" in front of goal. That is the only way, not the other way "- says Cruz. 
 * Is that him thinking like I was for a minute, that it must have been first fired from above, because of the wing damage? He also points to a smoke trail that should be visible for a BUK missile when there isn't (worth looking into) and he thinks they'd need more training than they had, while showing no sign of knowing how much training the relevant people had. Not too impressed with "Cruz." --Caustic Logic (talk) 09:49, 30 July 2014 (UTC)

I agree, it looks as if was hit with two different things. It is not possible that this happened at different times. The "two different things" can be explained by a mixture of armor-piercings and fragmentation ammo. (This is I suppose what a Su-25 would normally carry when attacking tank columns.) The real problem here is that I do not really know if these are exit or entry holes. I see some blast damage on the outside. The "cockpit roof / side panel" piece seems to show soot and slight buckling caused by an explosion on the outside. How long would it take for the frags to explode after passing through the aluminium skin?

Adam, could you create a stereoscopic image of the holes? Take two images of the same area, scale and cut them to exactly the same size. Two .jpg images would be good to flip-flop between. Maybe a flashing .gif for social media. -- Petri Krohn (talk) 13:23, 30 July 2014 (UTC)
 * Hm, maybe. Never tried that, they'll be at different angles, etc. Explain in e-mail. I'm slow today. --Caustic Logic (talk) 22:16, 30 July 2014 (UTC)


 * You need to be aware the panel has two different metals above and below the join line. The upper material is in immediate proximity to the windows and is much thicker and stronger - probably titanium. The lower material is Aluminum alloy - perhaps 80 - 160 thou thick. The splash pattern on the thicker/stronger material at the top is characteristic of high velocity impact in strong materials with subsequent cratering. The lower damage is characteristic of particles punching their way through thin metal with little or zero cratering effect.
 * --Charles Wood (talk) 14:20, 30 July 2014 (UTC)
 * Would this cratering be visible on the ingress or egress side? Or both? I would expect the the patterns to have radial structure, but they all seem to be tilted the same way. -- Petri Krohn (talk) 15:13, 30 July 2014 (UTC)

Shrapnel Fragment Size
I'm not sure this is the same warhead we'd be looking for, but likely - if the SA-11 has one type of warhead, this is probably it. Charles posted this photo of an unexploded SA-11 warhead, plus on another area (further detail yet, below). Where/when the photos are from, not sure, but they are the main feature in an August 2014 report by Richard M. Lloyd, available here. He thinks this is an SA-11 warhead, apparently the only one, and thus a clue to the MH17 case. He compares this with schematics to show it's the right size (you'd think he could have more definitive sources...) If the overall rocket is 5.5 meters long and .4 meters at the widest, the warhead area must be (about) 8 inches/ 20 cm in diameter at most, and thus maybe 12 inches/30 cm long? From there, I put the section with frags about to scale on the other image (it involves flipping one). At full resolution, it seems the big photo is about 150% life-size (the warhead measures about 12" by 18" on-screen), so the 3.75" exposed portion is about 2.5" real-life, if I scaled it right. If so, there are five rows of fragments, so they should be about a half-inch square (on-screen: app. .75" each) or about 13 mm - to holes in the hull measured as around 20 mm. This might be significantly off - it was a first try. And, considering the 3-D aspect, the final hole would be bigger than 13 mm, depending on the angle.--Caustic Logic (talk) 07:02, 10 August 2014 (UTC)


 * In impacts where there is significant surface depth, relative surface hardness, and high impact velocity you get a neat cratering effect that is similar to moon craters and splashes in water. The shape of the impactor is mostly lost and all that remains is a fairly symmetrical round impact crater.


 * It's not in the least unreasonable that a ~13mm irregular impactor into thick titanium produces an ~ 20mm circular crater - whether or not it actualy penetrates.


 * It's only when the surface is very thin and soft that cratering disappears and shape punching occurs. Of course this all depends on impact velocity. There will be a whole spectrum of shapes from simple indented punch marks to nice round craters blasting back in approx the direction the impactor came from (though mostly orthogonal to the surface) --Charles Wood (talk) 07:18, 10 August 2014 (UTC)


 * It indeed may have some similarities to cratering. (However, my understanding is that "classical" cratering assumes projectile velocity higher than material speed of sound (also deep material -no exit). Aluminum speed of sound is maybe 3000 m/s or more, above Mach 10 in the air. Most likely, projectile speed is less than that, so for impacted material it is likely "subsonic"--less than assumed in classical cratering).

--Chingachgook (talk) 12:55, 10 August 2014 (UTC)


 * That's simply not true. You can get 'cratering' in water by lobbing a stone into it at some tiny fraction of the speed of sound in water.


 * Cratering in metal is a complex interaction of solid, plastic, and liquid response with localised phase-change depending on impact energy. Essentially the impactor changes the local phase of the metal due energy release, and the bulk physical properties of the material and impactor may then produce a 'splash'. Nothing to do with the prior bulk material speed of sound. --Charles Wood (talk) 13:07, 10 August 2014 (UTC)


 * I agree with all that, still in deep material moving faster or slower than sound in material propagates makes a difference to those complicated details.
 * Craters on planets we observe are from supersonic events. Over here we most likely have it "subsonic".

--Chingachgook (talk) 13:36, 10 August 2014 (UTC)


 * Sounds reasonable to me. Also to note, if this device was used, I see little clumping happening. Little blocks and littler rods, all distinct, for consistent sized holes - 2 main sizes. Is that consistent with what we see? --Caustic Logic (talk) 08:20, 10 August 2014 (UTC)


 * My understanding of the warhead shrapnel is that it is in two layers and in each layer it tesselates two basic designs of different sizes - a larger butterfly pattern and a smaller diamond/cubish pattern (memory may be faulty so check) --Charles Wood (talk) 08:25, 10 August 2014 (UTC)

Interception Location
My best guess for interception location is at the intersection of the physical evidence heading east-west and the ephemeral trail heading North-West from that point.

I've done some searches on METARS at the time without useful result. My guess at this time it was a South-East wind - combined with a substantially West to East flight-track. Other METARS info welcome.

See image



--Charles Wood (talk) 11:12, 10 August 2014 (UTC)

That's an interesting graphic. It has way more dots than any I've seen. The Wall Street Journal's maps show less pieces, with whole areas blank, but the ones it does show, it says what they are. From this, it seems the nose came off first, and forward parts behind that. * The cockpit and lower nose cone came down like a bullet in that lonely middle dot. Higher drag but heavier forward fragments northwest, lighter forward fragments northwest of there. That's pretty clear - southeast means from that way, right? I've said northwest, meaning to specify to the northwest. These added dots give more shape and it seems a bit NNW. Hence, all debris is north of the flight path. The impact point will be back along that line, I don't think very far. FWIW I mapped it out and the intercept point seems to be no less than 25 km from the Snizhnoye firing spot, and I doubt it's more than 28 km out. But that's subjective, I don't know for sure how quickly it would fall apart. I know PA-103 popped like a balloon from the one little hole, but is it the same if there's a bunch of holes? Will it blow apart then, or just decompress more gently through all those holes and then tear apart over time? And how much time? How much does the decompression worsen the holes? Etc.
 * (not intercept-related: after the front came off, it looks like the rest of the plane crashed intact at Hrabove, with no real spread aside from what the fireball tossed up (that might have had a southward roll for whatever reason, and the wind to the NNW left it just south of the main impact - tail on the road, lighter stuff west) --Caustic Logic (talk) 11:38, 10 August 2014 (UTC)


 * METARS will be very useful. Too late for me to chase. If we know the wind direction we can start making sensible comments on the debris trails --Charles Wood (talk) 11:44, 10 August 2014 (UTC)
 * And/or we can measure the wind by how the stuff drifted. :) I also have as support apparent SE winds at Zuhres, but that was just after noon, and some kilometers west. We could also check the smoke plume video - U@War will know which direction that was filmed.--Caustic Logic (talk) 11:38, 10 August 2014 (UTC)


 * Hey! an ACLOS First? Assuming no spoilers we have the exact interception point within a few hundred metres! A Km at worst!
 * How to pubish momentous moment? --Charles Wood (talk) 12:00, 10 August 2014 (UTC)
 * Getting close anyway. At least you and I could agree on that - I like how that agreement has been working out. --Caustic Logic (talk) 12:46, 10 August 2014 (UTC)