DSB final report doubts: location of explosion
DSB calculated the location of the explosion as: X=0 Y=-2.0 meter Z= 3,7 meters
That means the missile exploded exactly two meters left of the nose and 3,7 meter above the nose. This position was calculated using software Split-X. The x,y,z axis are relative to the tip of the nose of the aircraft.
What DSB did to calculate was observe the shrapnel damage of the aircraft and then use all kind of parameters in Split-X (like speed of missile, azimuth and elevation) to determine the location.
Mind Splix-X software is not open source.
However when using photos of clear damage on the aircraft and use stringing technique, the altitudeof explosion above the nose is around 1,4 meters less then calculated by DSB. Three different persons did the stringing and all concluded the missile must have exploded at a height (z-axis) of about 2,3 to 2,5 meters high.
Almaz Antey calculated a different position of the missile relative to the nose than DSB took as most likely scenario. As the table below shows, Almaz Antey (model IIIc) calculated a height of 3.0 meters.
The image below was taken from the DSB report and shows the area (yellow circle) of the where the missile was when it exploded.
The document written by NLR shows what looks like a lower position of the warhead based on stringing. It appears the strings touches eachother at a low point than the yellow circle on the left picture.
Stringing
Now lets have a look at the damage to see if we can locate the location of the explosion using stringing. With stringing we draw lines from various holes or scratches (caused from ricochet) trying to find a point where all lines come together.
There are various close-up photos available on the internet showing clearly the trajectory of fragments.
Photo1: near the nose. Scratch running almost vertical
Photo 2: near the angle of attack sensor
Photo 3 jst behind the cockpit
This image shows the righthand side of the cockpit roof. The ricochets are clearly visible.
A couple of people including me used the damage on the photos to determine the location of the explosion.
Three different photos/impressions of the location.
My own effort to determine the location of the explosion can be seen in the image below. A high resolution image with the stringing can be viewed here.
Secondly a guy called Albert Lex did some stringing as well.
Compared the the DSB location the height of the explosion is lower.
Then an image made by a German person nicknamed Ole, The area where the green lines converge is more likely the area where the missile exploded.
The original photo of the cockpit in high resolution can be downloaded here.
Here is another photo showing the various holes. This thread at mh17.webtalk.ru has a lot of interesting info.
The image below was also taken from mh17.webtalk.ru
Conclusion
While the stringing using flat 2D photos and not having access to the actual debris is obviously not the most accurate way to measure, the difference between DSB calculated height and estimated height based on stringing is quite remarkable.
by
> While the stringing is obviously not the most accurate way to measure.
Stringing was the most accurate method in our and the DSB disposal. What made you to come up with the adverb ‘obviousely’?. The errors of the stringing method can be accurately estimated. The DSB, however, have decided to use a method that has an order of magnitude bigger specific error.
In additional to the stringing method there is another method, which gives the same explosion location – determining the detonation point by identifying where the fragment trajectories were tangential to the surface. Then one builds the tangential surfaces that also intersect in the same point.
This post is very welcomed. Thanks! However, I am a bit disappointing that, giving the far reaching implications of the detonation location, you’ve made a reasonably un-detailed short post. This is probably not an area you particularly enjoy or are fluent at.
What I meant to say is that stringing using photos (2D) has some risks on error. Also hard to judge what the cause of scratches is.
However I have my doubts about Split-X method as well
For historical justice, the first member of the public who, AFAIK, came up with detonation location was meovoto from mh17.webtalk.ru (it is called meovoto point on the forum). Then more people repeated his job and came up with the same location. Almaz Antey did that probably before, but it was not published. Now, you too intersected some lines and got to the same point.
A bit later I may post an estimation of the errors of the stringing method. I remember, when I did the estimation the errors was around 2 cm. This is on top of the additional errors associated with drawing lines. Judging by the fact that everyone who starts intersecting their own lines inevitably comes up with the same point, the latter error is also insignificant.
Dug up my early calculations for the non-straightness of the strings. My estimate was conservative (I was trying to err on the side of error exaggeration) and the resulting deviation from straitness over a course of trajectory was obtained to be 0.8 cm. The real-life error due to the lines not being straight would be even (say 2x-3x) smaller.
Now, given that, consider the decision of the DSB to switch from the method with specific error of less than a centimeter to the method with the specific error of meters, because “lines are not straight”.
What a sham.
The calculations (to be honest a more detailed description and pictures should be due, but anyway):
Fragment speed: 2400 m/s
Plane speed: 250 m/s
Air density (at 10 km): 0.4135 kg/m^3
Dynamic preassure (for Cx=1): P=1/2*0.4135*(2400)^2 = 1190880 Pa
Bowtie dimensions: 13mm x 13mm x 8.2mm
Iron density: ro=7874 kg/m^3
A bowtie mass (assuming square shape): m=d*d*d*ro = 0.013*0.013*0.0082*7874 = 0.0109117892 kg = 10.9117892g
A bowtie area: A=d*d
Find a bowtie decceleration
f==ma => a=f/m
f=A*P=d*d*P =>
a=(d*d*P)/(d*d*d*ro)=P/(d*ro)
For metal thickness of d=8.2 mm: a=1190880/(0.0082*7874)=18444 m/s^2
Time for a fragment to fly the length of its trajectory (take 5 m for for the length) t=5/2400=0.002 s=2ms
Speed change due to the decceleration during this time a*t=18444*0.002 = 36.888 m/s
Maximum possible trajectory angle change over this length (when the deceleration is perpendicular to the speed) = atan(250/2400)-atan(250/(2400-36.8)) = -0.0016044 rad = -0.09192 deg
Maximum error of stinging method due to trajectory curving: 5m * tan(-0.0016044) = -0.008 m = 0.8 cm
0.8cm guys!
Unless the type of warhead can be established by other means then stringing can only give us the position of the detonation relative to the y axis and z axis of the aircraft. This would be where a shrapnel fragment was traveling at an unknown speed in exactly the opposite direction (in both the vertical and horizontal planes) to the aircraft – it’s impact point was probably somewhere on the cockpit roof aft and just to the port side of the captain’s position. The x axis position of the detonation depends entirely upon the relative velocities of that shrapnel fragment and the aircraft. Additionally to establish the x axis position using stringing from other impact points on the aircraft the speed of the individual shrapnel fragments that created those impacts needs to be known. It can not be assumed that all the fragments burst from the warhead at the same average speed over various range distances as the aircraft moved into that expanding fragmentation cloud.
Red flag #1 – vague wording and unclear statements.
> Additionally to establish the x axis position using stringing from other impact points on the aircraft the speed of the individual shrapnel fragments that created those impacts needs to be known.
Can you, please, once and forever draw us a picture and explain how exactly “the speed of the individual shrapnel fragments” affects the calculated detonation point?
Previously you’ve been stubbornly refusing to make statements any clearer than these, draw pictures or to give the details.
http://tinyurl.com/jfvh2er
http://tinyurl.com/z466rxx
http://tinyurl.com/j2uten8
http://tinyurl.com/hb28t8z
I agree with the article. also A-A detonation test prove that to cause mh17 like damage 70kg warhead detonation must happen much closer to cockpit.
> A-A detonation test prove that to cause mh17 like damage 70kg warhead detonation must happen much closer to cockpit.
The damage in the AA test does look to be smaller than the observed on the Mh-17 wreckage. But one doesn’t need to move much closer to increase the damage. This is because the damage falls as a square of the distance (or even worse, see below). Let’s say that the observed damage in the AA test looks to be half of that on the Mh17. If we want to place the detonation point twice as close then we’ll need to actually halve the warhead mass if we want to match the damage observed on the Boeing wreckage. And, remember, we want to place the bomb up to three times as close. Therefore, one can conclude that we need a smaller warhead than that of a Buk, not larger. The only alternative to reducing the warhead size is believing the DSB detonation point, and after you’ve drown your own lines it’s pretty hard.
Here is more information on the detonation range versus explosive mass dependence. When it comes to explosions there are various scaling laws, that are often used to extrapolate results of small explosions onto the the explosions of bigger sizes. One of the scaling laws is the Hopkinson-Cranz Scaling Law. Check out the applet here:
http://www.un.org/disarmament/un-saferguard/hopkinson-cranz/
In the applet you can enter the explosive weights of 1 kg and 8 kg, and calculate the safe ranges for the two blasts. You’ll see that even if you’ve increased the explosion weight 8-fold, the safe distance (that is the distance of a constant and small damage) only doubled. That should give you the feel for the very sharp dependence of the damage on distance.
Off topic. Sotilaspassi! My article on Winter War has been published in Fort-Russ. It would be nice Finnish opponents, including fierce opponents. Any feedback is welcome. http://www.fort-russ.com/2016/04/finland-winter-war-of-39-40.html
What I miss in all the graphics is the orientation of the missile itself.
Because this is important to specify the direction in which the missile fragments flew (may they have reached the left intake ring or not).
I am working on a blogpost about the azimuth (postition on the horizontal field) en elevation (position of missile in vertical field)
>>DSB calculated the location of the explosion as: X=0 Y=-2.0 meter Z= 3,7 meters
No. DSB is not say about exact location. We know that DSB located blast point in volume 1 m3. But we dont know center of this volume. Also we dont know exact result microphine calculation. In DSB report we can see only coordinates from other TNO, AA and etc.
This picture is marked “Not to scale”. I suppose location of blast point on the picture did’t equal DSB results.
http://s014.radikal.ru/i328/1510/66/413dcae3acaf.png
As I understand, DSB dont want to open results now.
The best picture id DSB report for search blast point is it:
http://s05.radikal.ru/i178/1604/64/2ea5989fdd1e.png
I see 2 focus point. It is very right. Heavy(slow) and light(fast) fragmenys of warhead must get diffrent blust position, because Boing have non zero speed.
http://unit0.livejournal.com/1120.html
Also about doubts. In this post, I calculated maximum exaction angle for 9N314M1 warhead using IL86 test data. The result is 115.5 deegrees as minimum. The maximum exaction angle is very important for search of start point by modeling. And numbers from DSB report and AA data is very diffrent 112 and 126.
unit0, your English is a bit hard to understand. For example the term “exaction angle” is of your own creation, probably meaning one the limiting angles of the cones confining the trajectories of the pre-formed fragments.
> And numbers from DSB report and AA data is very diffrent 112 and 126.
The DSB have misrepresented the data about shrapnel cones supplied by AA. The data supplied by AA would not allow even with the use of an unknown matching method to move the detonation point away, because the simulated damage strip would be too wide. Therefore the DSB had to shrink the cone. You can almost see a nontrivial maneuvering job the DSB had to do to come up with a justification for a BUK warhead. And, indeed, when AA placed the warhead into the detonation point found by the DSB the resulting observed damage strip was too wide. For example, in the AA test there were holes before the most forward structure rib shown here on Photo 1, or holes behind the cockpit door. Compare the areas covered by hits http://savepic.ru/8334698.png
As to the calculation of the confining angles, I personally would not trust any calculations, whether Split-X or yours, when a simple test explosion can be performed. One of the reason is that the Buk warhead may contain a primer booster, which is not taken into account by calculations, or may not. The best we can do here is either try to determine the cones from videos, or just trust AA.
The piece 2, the one with the angle of attack sensor, is slightly off horizontal. I don’t know whether you took this into consideration or not. The intersection point would have been just a bit lower if the misalignment is factored in.
> The image below was also taken from mh17.webtalk.ru
It was actually me who was the original author of that one. The model was a bit wrong there, but still allowing to see the point. I was about to fix the model when AA came forward with the same location. After that the point became almost official, so I abandoned the work.