PUBLIC Forums => Public Forums => Forensics => Topic started by: Brandon Bertolli on November 18, 2007, 09:10:33 AM
-
Can you determine the angle at which a windshield was perforated if you don't have any evidence on either side of the glass? What I mean is, if I perforate a windshield and bring you just the windshield can you tell me what the angle of incidence of bullet to glass was?
I understand that the bullet changes trajectory upon impact because of the slope of the glass relative to the nose of the projectile. I've been told that a horizontally-fired bullet that hits the exterior of a sloped windshield will have a minor trajectory change towards the ground and that a horizontally-fired bullet that hits the interior of a windshield will have a trajectory change towards the sky.
So my second question is:
Is there enough thickness of windshield glass and adequate techniques available to determine what the trajectory change was, in circumstances where the round is known to be intact, but the round has not been recovered and nothing else about the shooting is known?
Thanks again for the knowledge.
-
Good morning, Brandon,
This is an interesting question. Most of the work involved in the calculations for angle of incidence depend upon the nature of the substrate being hard and unyielding. Although empirical testing is always beneficial, the calculations in this regard appear to be analogous to those for bloodstain pattern analysis.
However, upon closer inspection of what is actually occurring during bullet-windshield interaction, I think that one must be very cautious when attempting to estimate the vertical angle element. Obviously, when compared to bloodstain pattern interpretation, the blood is yielding whilst the substrate is typically not. Conversely, whilst the bullet is relatively hard, it may deform and/or fragment, not to mention be deflected to some degree (as you already mentioned). In addition, the windshield itself is not 'unyielding' - the laminate flexs quite dramatically during impact and oscillates, and can also break irregularly from shot to shot.
On page 180 of Ed Hueske's book, "Practical Analysis and Reconstruction of Shooting Incidents" (part of the Practical Aspects of Criminal and Forensic Investigations Series), Ed writes:
"The vertical angle cannot be reliably determined from the length and width measurements of bullet holes through windshields because of bullet deformation and fragmentation and the irregular breakage of windshield glass. In the one test conducted by the author with 40 caliber Federal Tactical rounds, no reproducible results could be obtained and none of the vertical angles of impact calculated from bullet hole width-to-length ratios were within fewer than 15 degrees of the actual."
Personally, I am always in favor of actually doing some testing of my own - and this sounds like a great day at the range! If you know the firearm and ammunition believed to be responsible for the shooting incident, I would go ahead and do some testing - but being mindful of the potential limitations. If you do decide to go this route, please be sure to share your findings!
Best wishes and warm regards to all,
-G.
-
Geoff, thanks for such a detailed answer. It looks like there is no easy way to determine this, then. At least not with the glass alone.
-
Hey Brandon.
Agree with Geoff. Empirical testing is always best.
Thought I'd throw in 2 cents here also.
For the azimuth angle, the long axis of the ellipse is good, but only as viewed from directly above the impact, with some relevant reference line in view.
I put little to no resolving power in the elliptical shape of the hole for the vertical component (for common angles of impact), however. As Geoff said, because glass is so frangible, the proportionality of small d to big D ellipse is not consistently representative to any degree of certainty other than saying it is a relatively flat trajectory (assuming we are dealing with the common approach angles we usually see for windshields). Because of this lack of resolving power, the vert component usually doesn't help to position a firearm 5 vs 50 ft away.
As for the deflection issue, I have done several well documented experiments, as well as multiple less intense studies at my shooting recon classes involving deflection, and for handgun bullets going into the glass at the common windshield angles, downward deflection on the order of 1 to 10 degrees is common. This translates to about 1 to 6 inches downward at the common range of a shootee from the windshield. Slight right to left seems unpredictable. Shoot outward, and indeed the deflection changes direction. Important to note here that the distance from the windshield to target may now increase so observed deflection can be significantly greater. However, be cautious in that once you get extremely shallow, or change velocities, the deflections can be different... This general rule applies to common handgun bullets and velocities.
The "bloodstain" trig functions do give good approximations of bullet impact angle in materials like drywall where the front half of the perforation is fairly intact... I have done known angle verifications of this... a minimum of +- 5 degrees seems appropriate, like with many other trajectories.
Sorry to be so long winded.
mh
-
Thanks, Michael, that is great information. I do appreciate everybody's effort here in educating me on this issue.
I am going to have to read up on blood stain trig functions because I am ignorant in that regard. But generally, from what I hear so far, it seems to me that even if you had a windshield perforation and a fabric breach in a car seat, you couldn't work out the angle of incidence of the bullet to the glass, because of the unknown deviation in trajectory from the time the bullet interacted with the glass to the time it hit the seat or whatever target it hit within the vehicle. Is that a fair statement to make, or can you specify a range of angles if you know roughly what the velocity of the projectile was and you recovered it intact?
-
From my experience, I would suggest a minimum of +- 5 degrees if you have a subsequent impact in the seat. Depending on how shallow your impact is to the glass, maybe even +- 10 degrees. Remember the tendency with handgun bullets at common velocities will give slight downward deflection (out to in)...
The azimuth component is usually more informative becasue there are a wider range of realistic possibilities in the azimuth plane (ie, all around the car), whereas the vertical components (which may normally be used to establish gun location out from the vehicle) of such shots are usually pretty "flat" / horizonatal. The vert component's accuracy in glass isn't suffient to get accurate info becasue the +-5 or 10 degrees opens up the range of possibilities so wide. This isn't just glass though, as establishing gun from object distances with flat trajectories is difficult for these same reasons... The steeper the shot, the easier it is.
Tough concepts to type. I hope that makes sense.
Take Geoff's advise though...
Shoot stuff under controlled conditions in order to give yourself a proper education in these areas.
mgh
-
"...In addition, the windshield itself is not 'unyielding' - the laminate flexs quite dramatically during impact and oscillates, and can also break irregularly from shot to shot..."
Some readers may not know of US' windshield requirements.
European windshields (last I checked) are not laminated—but tempered. (Laminated means having a plastic ply between two glass sufaces). Side windows and the rear window are tempered.
Also, not all windshields are alike: modern bonded windshields are much thinner and lighter than their predecessors. (And why we see so many windshields being penetrated—and passengers struck—by workaday construction debris tossed up from roadways today.)
I would shoot tests with the windshield bonded in place, as the windshield's strength is otherwise compromised.
Sometime, watch the reflection of a modern windshield outside while someone cleans the inside surface. You can actually see the glass flexing at each finger: scary and impressive at the same time. :o
-
@ All
That's interesting information regarding laminated vs tempered glass.
I have a case in my manuscript (not yet a published book) that involves a bullet that perforated the side window of a VW Golf and continued on a downward trajectory and struck the driver on the left thigh. The bullet remains in the patient, but the wounds and the radiological appearances of the bullet are helpful in verifying the patient's statement in terms of the circumstances of the shooting. I'll find one of the radiographs and post it here. The bullet has a definite flattened nose and I am wondering if the angle of deformity relative to the long axis of the bullet is a close enough match to the angle of glass relative to the barrel at the time of the shooting.
-
Adding on to what Bob said, not all laminated windshield are the same in construction.
Some more expensive cars (BMW, Mercedes, Lexus, etc.) have multiple layers of laminate and glass, ie, more than just one sheet of plastic as is most common. I have heard of but not observed 5x layers.
Unfortunately, I have never had a BMW shot in ABQ. I'm still waiting.
Also, refer to "Shooting Incident Reconstruction" by my Padre, Luke, for info about the "flat" spot deformation vs. angle of impact. Particlularly in cases when rico / departure angle is not reflective of the incident angle (as in most cases), this damage is the only good way to estimate in-shoot / incident angle. In short, yes, it is a good approximation of the angle of impact, assuming significant subsequesnt damage was not done to the projectile. A nice soft leg should be a great decelerator as opposed to steel structural material. This is also one of the many pratical demonstrations at shooting recon classes.
I'd be curious though how to insure with a radiograph you are getting a propper orientation view to assess this angle on the bullet still in a leg.
Look forward to the book!
mgh
-
Ladies and gents, I have located the radiographs:
No 1: AP Pelvis
(http://i55.photobucket.com/albums/g154/Odd_Job/700X-rayPelvis.jpg)
No 2: Horizontal beam R thigh
(http://i55.photobucket.com/albums/g154/Odd_Job/712X-rayHBL.jpg)
It was only a matter of good fortune that the standard radiographic planes resulted in radiographs being obtained roughly parallel and perpendicular to the long axis of the bullet. I don't think you could measure the angle within 1 degree accuracy but it is my claim that in this case the radiographs might be helpful in determining the angle of incidence.
Unfortunately, because of the density of lead, if you don't have the X-ray beam tangential to the impact surface of the bullet, you can come to misleading conclusions about the state of the bullet. You can emulate the beam projection dynamics by drilling a small hole through the length of a bullet and inserting a glass rod through it. You can then get an idea of the radiographic contour by shining an ordinary flashlight on the bullet and observing the shadow on a piece of card.
I guess you could also ray-trace it with 3D modelling software which gives you the advantage of being able to modify the geometry of the impact deformity on the fly.
Anyway, I fully appreciate the problems of documenting an impact surface radiologically, because even if you have the base of the core relatively flat on the radiograph, you are still left with variables associated with the rotation of the bullet about its long axis, relative to the beam. If this case went to court, I would probably suggest multiple projections, covering a 20 degree arc (10 degrees posterior and 10 degrees anterior angulation relative the the original horizontal beam lateral). If I get the time this week, I'll do some ray-traces to illustrate what I mean.
By the way, here are the images I have used to illustrate the possible trajectory of the bullet in this case. I didn't create the model of the Golf, but I did add some furnishings inside the vehicle because it was provided on-line as a mere shell. There are several small discrepancies based on the fact that this was not the exact model of the Golf in question and also the colour is not accurate. I was not allowed to see the vehicle, but the patient offered detailed information about the circumstances of the shooting.
(http://i55.photobucket.com/albums/g154/Odd_Job/729GolfFront-1.jpg)
(http://i55.photobucket.com/albums/g154/Odd_Job/730GolfFront-1.jpg)
I have wound photographs also, but I am holding those back because they are of a sensitive nature. The man's scrotum was perforated and the photographs include his penis. You'll have to trust me when I say the entrance wound was consistent with a round that had already expanded or was otherwise deformed prior to impact.
On the radiographs you can see the paperclips that we have used to mark the skin breaches in this case. The position of those paperclips radiologically can help prove what the position of the patient was, at the time of the shooting. I'll give you a hint: on the AP view the paperclips don't line up because the patient was X-rayed supine.
-
I must add also, that it is not clear who the target in this shooting was, because the shot was fired from the passenger side of the vehicle (which is right hand drive, being in SA) and the man's girlfriend was seated in the passenger seat at the time.
-
Hmmm. Maybe the intended target was struck. :-\
To me, your radiographs indicate a lead (non-jacketed) bullet.
I was just about to comment that I've seen lead bullets that have penetrated plate glass have an "affect" similar to this image I found on the Web:
(http://images.jupiterimages.com/common/detail/92/73/23057392.jpg)
(The image is unrelated, and you have to ignore the diagonal line.)
That is, a mirror-flat bright surface upon impact (the round area), surrounded by many other fractures that are also very highly reflective.
If anyone needs an AFTE study subject, it would be interesting to see the effects of plate glass, laminated glass, and tempered glass on trajectory through each. I'd suggest using non-jacketed lead bullets only to see if that mirrored-affect is different for the three different glass types.
-
Sir, the round is definitely jacketed, but the size of the image I first posted is not sufficient to show it. Here is a magnified view, showing the low density metal of a jacket fragment:
(http://i55.photobucket.com/albums/g154/Odd_Job/719Closeupbulletjacketarrow.jpg)
The fine horizontal lines across the entire image are caused by a stationary grid designed to improve image quality by reducing secondary scatter reaching the film. The white band at the bottom is cortical bone of the femur. No bones were struck in this case.
I agree on the glass research. Some people may already be involved in that, and I eagerly await their findings.
-
Sir, the round is definitely jacketed, but the size of the image I first posted is not sufficient to show it.
As usual, I'm still puzzled; however, please substitute just sayin' for my prior indicate. ??? Hijacking of the thread was not intended by me :-X
-
It's not hijacking, your questions and comments are most welcome. You guys have always helped me out with info, and I'll do the same if I can.
The thing that differentiates lead fragments or cores from non-lead jacketing is the radiological density.
An easy test for determining whether something is lead or not, is to try to see bony structures or other anatomy 'through' the metal. If it is a completely white (radio-opaque) fragment then you can bet you have found a lead fragment, or in some cases a large steel insert. This all assumes that the radiographic exposure is not too light overall.
But certainly, in all cases, if you can see anatomy 'through' a metal fragment then that fragment is almost certainly jacketing. It could be steel or copper-based, it doesn't matter. It is essentially a thin strip of low-density metal which has a distinctive appearance radiologically.
The best example to illustrate this is a core-jacket separation. I will locate one of those radiographs from my library and post it here.
-
Okay, here is the radiograph:
(http://i55.photobucket.com/albums/g154/Odd_Job/Figure7-1.jpg)
That is a classic core-jacket separation. This was an FMJ of unknown calibre that entered on the right lateral aspect of the pelvis. No bones were hit, there were no fractures.
The core is a completely radio-opaque white shadow (double white arrows) while the jacket has lower density (it is more grey): enough to see the femur bone behind it. If the core was overlying the femur, you would not be able to see any part of the femur bone behind or in front of it.
-
Even if these are not the best X rays, I would suggest not making judgements about impact angle from these...
Even an actual exam of the projectile must be done cautiously...
Doesn't look like enough info to me from these images to make good impact angle judgement.
Make 'em cut it out... Hehe.
mgh
-
I wish they would cut more of them out (but that's just me trying to get a complete documentation of the projectile and the damage it caused). There were two cases where I really wanted the projectile but in each case it was not clinically possible. Scott Doyle and a gentleman by the name of Emil Hamza helped me ID one of the projectiles radiologically and in the other case, the projectile remains unknown.
-
Brandon,
When you say "ID one of the projectiles radiologically", what do you mean?
-
Michael, it was a case that had baffling appearances to me, because there were multiple lead fragments and a fully intact symmetrical high density 'thing' embedded in the tibia.
We don't see many cases like that in the civilian setting, and that is the only one I have seen, period. To cut a long story short I wanted to know what ammunition this could have been, and it was at that point that the fine members of this board advised via Scott Doyle that the strange item was most likely a steel insert. It then remained for the ammunition to be identified by the unusual contour of the insert. Emil Hamza suggested a possible match: 7.62 x 25 Tombak-plated steel jacket steel core Tokarev of the Czech variety. His photographs were very similar to my radiograph in terms of the contour of the insert (which could not be removed from the patient and was not seen at surgery).
I managed to source some of those bullets here in the UK, extract the insert and X-ray it here at an equivalent angle as the one lodged in the patient. The contours were an exact match and no other insert that I have investigated so far has those same contours. So I am quite confident that this was a Tokarev insert.
It took a while to arrive at that conclusion: the guy was shot in Johannesburg in 2002 and I only came up with the answer towards the beginning of 2005 here in London. It is one of 150 cases I documented in four months in Johannesburg. These are the cases I will reference in my book (which is on ice at the moment because of work commitments).
-
Okay, here is the radiograph:
([url]http://i55.photobucket.com/albums/g154/Odd_Job/Figure7-1.jpg[/url])
That is a classic core-jacket separation. This was an FMJ of unknown calibre that entered on the right lateral aspect of the pelvis. No bones were hit, there were no fractures.
The core is a completely radio-opaque white shadow (double white arrows) while the jacket has lower density (it is more grey): enough to see the femur bone behind it. If the core was overlying the femur, you would not be able to see any part of the femur bone behind or in front of it.
That is a good example.
A few observations from my prior caliber x-ray determinations:
1) When I've put known caliber fired bullets on the x-ray platen for scale, often I've seen the jacket surrounding the core like a halo. (Requires a very favorable exposure).
2) As to "unknown caliber": an x-rayed bullet can only appear larger than is measured directly from the x-ray. IOW, one can assert that the fired bullet's actual caliber is no greater than what can be directly measured on the x-ray itself. In the above x-ray, that would be a difficult determination without additional x-rays to align the core axis (if available) to the platen.
3) A bullet's caliber can be determined using the "Peter's Method", previously described in the AFTE Journal. It requires several x-rays of the victim to help orient the evidence bullet, "coning down"—and the use of a "tomographic phantom". (Which should be amongst the x-ray lab's equipment.)
The topic can be argued endlessly, but I find x-ray bullet caliber determinations fascinating. :)
Once, a prosecutor asked me to make a caliber determination in the hallway outside of courtroom!
The victim—who was to testify—was in a wheelchair due to the first shooting, and had just suffered a subsequent shooting ( :P ). The bullet was caught just under the skin. It was quite the experience to palpate a bullet caught in a living shooting victim. :o
To the prosecutor's question, "Was it consistent with 38 caliber?", I could say "Yes".
The judge threw out my testimony because there had been no "Discovery". (Although the defense attorney was standing beside me at the time of the examination.) ??? :-\
-
1) When I've put known caliber fired bullets on the x-ray platen for scale, often I've seen the jacket surrounding the core like a halo. (Requires a very favorable exposure).
I've seen this also, usually at the base of bullets that have an exposed lead core. Beware, there is a false halo that can be caused by movement artefact (bullet is alongside a pulsatile vessel or organ) or also geometric penumbra from a large focal spot on the X-ray anode (older equipment, particularly mobile units).
2) As to "unknown caliber": an x-rayed bullet can only appear larger than is measured directly from the x-ray. IOW, one can assert that the fired bullet's actual caliber is no greater than what can be directly measured on the x-ray itself. In the above x-ray, that would be a difficult determination without additional x-rays to align the core axis (if available) to the platen.
One factor that will become increasingly influential is the digital age which is resulting in an increase in computed and digital radiography systems worldwide. Even the UK Forensic Radiography Response Team's equipment is almost entirely digital (screening and plain film): only the dentals are wet developed and solid state detectors attached to laptops are going to render dental films obsolete in a few years.
Anyway the problem with digital is the scale. You can now have an output image that is artificially smaller than the actual latent image on whatever detector is being used. If the calibration of that detector or the software that manipulates it is not right, you have a radiographic contour which cannot be described relative to the size of the bullet in situ. The same applies to images acquired on an image intensifier, because there is an intermediate step in the image chain where electrons are focused, and that means you've lost your relative size increase of the latent image.
3) A bullet's caliber can be determined using the "Peter's Method", previously described in the AFTE Journal. It requires several x-rays of the victim to help orient the evidence bullet, "coning down"—and the use of a "tomographic phantom". (Which should be amongst the x-ray lab's equipment.)
Measuring bullet calibres radiographically is kind of a sore point with me. I am aware that some members have techniques that they employ to make these determinations and I have also read up on the Peters method. These methods rely on the comparison of the bullet in situ to an approximate match or control bullet which is X-rayed under the same 'magnification' as the bullet in situ.
I have serious reservations about these methods.
In fact the only case I have read where I have any faith in the outcome of the calibre determination involves shotgun pellets where the suspect pellet lodged in the orbit was known to be one of only two or three possible sizes. I will have to find that paper for you, but the bottom line was that the pellet in situ was seen to be spherical on two projections and samples of the known sizes of pellets involved (two different hunters) when X-rayed at the same or similar magnification as the pellet in situ, revealed that only one size pellet was a possible match.
My beef with all the other methods is that there are technical factors and variables to do with deformity of the bullet that make it impossible to eliminate or compensate for differences between the radiological contour of the control bullet vs that of the one in situ. I find it unacceptable when the calibre determination is made in instances where there isn't a shortlist of suspected calibres with sufficient size differences between them to render such inaccuracies inconsequential.
Some of you may be interested to know (and some of you already do know) that I spent quite some time trying to develop a technique to determine the calibre of a retained bullet by radiological means. My method did not rely on a control bullet: it was an absolute measurement with all the variables leading to its magnification determined by CT scanning of the 'subject' in the same position as it was X-rayed.
Under the best conditions, using a very compliant and motionless deli sausage, I was able to get down to within 1mm of determining the size of the retained bullet, based on two sets of measurements in two planes at 90 degrees to one another. You might consider this to be accurate, but I found the accuracy of the measurement to be heavily influenced by where the person chose to physically or electronically determine that one edge of the bearing surface sits, as a starting or ending point for the measurement. When I factored in the variables associated with a live patient and the nature of the tissues involved and the fact that two views of the bearing surface at 90 degrees to one another might not be as easy to obtain in a patient as a sausage, I realised that despite my best efforts, my method would never be accurate enough to be used in as serious a matter as determining a man's innocence or guilt. I must add that these were bullets fired in a water tank at a ballistics lab and therefore represent bullets that are in very good condition as far as fired bullets go.
I really did try, people. I got medical physicists, equipment technical sheets, the cooperation of my colleagues and I even had a purpose-built jig for the X-ray cassettes and the sausage.
But I have no confidence in it. And my insight into the various pitfalls associated with X-ray imaging of this nature leaves me with no confidence in any technique I have read about thus far, with the exception of the one case I mentioned earlier (involving the shortlist of only two or three pellet sizes).
Anyway, I apologise for the ramble. I have issues with this calibre business, that's all.
-
"...Anyway the problem with digital is the scale. You can now have an output image that is artificially smaller than the actual latent image on whatever detector is being used...The same applies to images acquired on an image intensifier..."
I may be selling U.S. x-ray technology short, but a hospital x-ray of my ankle last summer seemed to require a wait of several minutes in order to view it. If the day arrives that a "wet" x-ray becomes unavailable, then there will be issues if a bullet can be imaged as smaller.
"...My beef with all the other methods is that there are technical factors and variables to do with deformity of the bullet that make it impossible to eliminate or compensate for differences between the radiological contour of the control bullet vs that of the one in situ. I find it unacceptable when the calibre determination is made in instances where there isn't a shortlist of suspected calibres with sufficient size differences between them to render such inaccuracies inconsequential.
Bullet deformity will be a problem. Most cases presented to me involved a handgun. Handgun bullet x-ray profiles are more revealing as to caliber.
My very first case was a literal riot!
Sergeant Fred Pelny of the Metro-Dade Police Department was wounded during a riot, and the firearm presented was a 44 Spl Charter Arms revolver. I was also presented with his ample x-ray with the question, "Can you determine what caliber this is?"
I could see a large caliber bullet—in profile—with a large, flat, meplat. O0
Determining the "largeness" of the caliber using the Peters Method, I positioned comparable caliber "knowns" and included other calibers known for their meplats.
Whilst a very slight difference in profile diameter could be determined by overlapping the Pelny x-ray with the knowns*, only the defense attorney was annoyed at the sight of an obvious 44 Spl bullet in Sergeant Pelny's x-ray. The judge allowed the methodology and the jury subsequently concurred with my findings.
*taken at the same determined height from the platen.
I'm an advocate of the Peters Method, but wouldn't attempt a caliber determination of deformed bullets in general and would be averse to strictly-rifle calibers.
"...Under the best conditions, using a very compliant and motionless deli sausage, I was able to get down to within 1mm of determining the size of the retained bullet, based on two sets of measurements in two planes at 90 degrees to one another. You might consider this to be accurate...Anyway, I apologise for the ramble. I have issues with this calibre business, that's all.
One millimeter isn't good enough, but combined with the bullet profile, I would consider handgun bullet caliber determination in "wet" x-rays to be a valuable endeavour.
Since several x-rays are needed for positioning the unknown bullet profile ever closer to the platen in the Peters Method, a pulsatile image should be detectable and rejected from consideration.
(A handy piece of information, though. I've developed a small pulsatile condition in the vision of one eye this year :P —and was advised that it was harmless 8) ) .
-
I'm an advocate of the Peters Method, but wouldn't attempt a caliber determination of deformed bullets in general
And that is one of the problems: how do you determine radiologically that a bullet is not deformed?
More than 90% of the ballistic injuries I have seen have been from handguns, and the imaging has been on wet films. My test with the sausage was also on wet films with CT used to measure the bullet-to-film distance in two planes. I assert that no claim can be made that the bullet is not deformed, because such deformity may be of the type that puts the bullet 'out-of-round' and it will only be apparent if an X-ray is done with the beam aimed exactly parallel and central to the long axis of the bullet. I haven't seen this done in any verifiable way (even by an accidental shot) and I can tell you that I have seen one or two gunshot wounds in my time (young though I am).
Have a look at this one, from my files:
(http://i55.photobucket.com/albums/g154/Odd_Job/Figure7-2.jpg)
This was an adult male struck in the head by a bullet. According to his statement, he didn't see the shooter and didn't know where the bullet came from.
The bullet looks like it isn't deformed, on both views which were taken at 90 degrees to one another. The main difference between the two views is magnification of the bullet, because the area of interest has a different bullet-to-film distance from one radiographic view to the next.
Yet here is the bullet, recovered by the trauma surgeon under local anaesthetic:
(http://i55.photobucket.com/albums/g154/Odd_Job/Figure7b.jpg)
Black arrow indicates the surface that was influenced by an intermediate target (highest probability) or skull bone (very low probability). My guess is that it was a masonry ricochet.
This is one of many cases that causes me to be concerned about calibre measurements in general. Clearly it has a minimum and maximum calibre measurement depending on the orientation of the bullet towards the beam at the time. So which profile are we seeing in each radiograph?
We don't know! It could be a profile that I don't even have an equivalent photographic representation of.
Then there are variables that we haven't factored in:
1) The difference between the radiographic outline of the bullet in situ vs the control bullet, in terms of edge degradation from scattered radiation within the body and also composite edging from the anatomy itself.
2) Variables to do with jacket thicknesses and metallic composition and the effect of this on the attenuation of the X-ray beam at the bearing surface.
3) (This one is most serious): the means by which one pair of edge intervals is measured and/or compared to the other. If you are doing it with a digital caliper, you are just eyeballing it. I can produce a radiograph with multiple exposures of the same bullet with all the conditions constant except the radiographic exposure itself, and it is my claim that in a controlled experiment you would all come up with different measurements of those bullets. When I say different, I mean not just from one exposure to the next, but also amongst different viewers measuring the same radiographic exposure. The reason for this is that unlike an actual bullet, the radiographic shadow has no physical 'stop' against which you can adjust the calipers. You have to apply a three dimensional tool to measure a purely two-dimensional interval. I have tried it also and didn't get consistent results.
Even if you employ software to do the measuring, you can be confounded by a bearing surface edge that has a grey scale that drifts from white to the black or grey of the surrounding anatomy. Sounds simple to measure, but just you try it! I have a high resolution transparency scanner here and it is very difficult to determine where the bearing surface ends and the surrounding anatomy begins. I tried the calipers and the scanner in the sausage experiment and although the scanner is good for eliminating dexterity and 3D vs 2D measuring, it is still no picnic to measure those intervals consistently. The difference between the resolution of two bearing surfaces may even be appreciated in a single exposure. All it takes is a difference in composite anatomical density to render that difference radiographically.