Question about Values in Shell_Table.lau

[Curly]

There seems to be some question in regard to the accuracy of the gau-8 and the values in Shell table. Modders have been lowering the value of Da0 to produce tighter dispersion patterns without knowing what the value Da0 means. I became a curious myself and looked into it, only to come up with more questions not only about the mysterious figure Da0 but also about how dispersion is handled in DCS.

 

 

Sources confirm that Gau-8 is rated 5 mil, 80%. Which according to some means it will put 80 percent of it’s bullets through a 5 mil circle at 4000 feet. Or for the metrically inclined, a 6 meter circle at 1,200 meters. After doing some additional reading, I think there is some misunderstanding of the rating of “5 mil, 80%”.

 

If we look “Predicted Effect of Projectile Dispersion on Target Hit Probabilities and Dispersion Zones For the 25-MM Gun of the Bradley Fighting Vehicle” , http://www.dtic.mil/dtic/tr/fulltext/u2/a193618.pdf

We can see the way this rating is calculated. The diameter of the dispersion zone in mils = 2 * the standard deviation of the dispersion times the C value of the Chi Square for 2 degrees of freedom for a given probability.

 

Or D= 2 * sigma Sqrt C

 

since we have D for GAU-8 (5 mils) and we can look up C in any Chi Square table. we get

 

5 = 2*sigma Sqrt 3.22

 

Which means sigma = 1.3932.

 

With a dispersion of 5 mils; 80% of the rounds will hit within a circle of 1.392 mils.

 

That's a much tighter pattern than a 6 meter pattern at 1,200 meters. Using that same linearization as before, we are looking at a 2 meter spread at 1,200 meters.

 

This pattern seems legitimate when you look at the results of the GAU-8 effectiveness study which saw a 49% direct hit from 790 meters, using a manual sight. http://www.dtic.mil/dtic/tr/fulltext/u2/a522397.pdf

 

What got all this started was trying to figure out what Da0 is. And how to manipulate it to render a result that is accurate as possible. Da0 in the shell table for all the GAU-8 rounds is listed as .0017. Which doesn’t look like it would be a standard deviation for a gaussian distribution of shells around the mean aim point. This lead me to the question if we are using a gaussian distribution to calculate bullet dispersion?

 

Further looking at some other equations leads me to think that shell distribution may be linear and based on the geometry of cones. Diameter of circle in cone = 2 h tan(the angle of the cone in radians), So perhaps da0= the size of the angle of the cone in radians?

 

4.080= 2 * 1200tan(.0017)

Which gives us an area of 13.07 meter. Meaning we’re evenly distributing all the rounds from a burst in a 13 meter or 42 foot cone. Which empirically seems to match the current in game behavior.

 

Given the small area of the targets and large dispersion radius it’s no wonder that hit rates are so low. If Da0 is what I assume, there simply to much miss area in the pipper for the gun to behave realistically.

 

if we want a 1.392 meter area at a distance of 1200m Da0 should be .000000307257.

 

I was wondering if we could get some feedback from the devs about the nature of Da0, what its, how it works and how bullet dispersion is calculated. Is it done linearly or is a method of normal distribution used?

Thanks for your time.

Edited March 30, 2017 by Curly
Corrected typo; replaced yards with feet.

[Dauntless]

My head hurts.

 

:pilotfly:

[Khamul]

Sources confirm that Gau-8 is rated 5 mil, 80%. Which according to some means it will put 80 percent of it’s bullets through a 5 mil circle at 4000 yards. Or for the metrically inclined, a 6 meter circle at 1,200 meters.

I think 4000 yards are not 1200 meters. It seams 3657m.

 

EDIT: now I see, you meant feet. From Wiki: " According to Dennis R. Jenkins, author of Fairchild-Republic A/OA-10 Warthog, 5 milliradians equates to a 40 feet (12 meters) diameter circle at the weapon's design range of 4,000 feet (1,200 m).[15] According to the milliradian system in use in the U.S. however, an angle of 1 milliradian equates to 1 meter (3.3 ft) diameter at 1,000 m (3,300 ft);[16][17] meaning 5 milliradians equates to 20 feet (6 m) circle at 4,000 feet (1,200 m). By comparison, the M61 has an 8-milliradian dispersion"

 

Sorry

Edited March 30, 2017 by Khamul

[al531246]

[Curly]

My head hurts.

 

:pilotfly:

 

 

I’ll try to clarify a bit. It’s incorrect to draw the conclusion that GAU-8 will fire 80% of it’s rounds through a 6 meter circle at 1,200m, based on the rating of 5 mil 80%. The equation and application in “Predicted Effect of Projectile Dispersion “ shows that the area and percentage in the rating 5 mil 80% is based on the variance from the mean aim point. Since we have the guns rating we can figure out the standard deviation. In the case of the GAU-8 the standard deviation of where the rounds impact around the aim point is 1.3932 mils. This is a much tighter grouping than 80% of the rounds through a 6 meter circle. That grouping would actually be 17.94 mils 80%, as the standard deviation would equal 5.

 

The shell table value for Da0 is .0017. Which could be the standard deviation based on the misunderstood nature of the rating 5 mils 80%. If it were mis-typed in error by 1 decimal, or obfuscated by a magnitude of 10 per ED’s contract with the DoD. Though I don't think this is the case because the grouping shrinks when Da0 gets smaller. So It leads us back to the question what is Da0?

 

 

I have done some informal testing with Da0 set to .000000307257 and the results seem to be more in line with results in the damage assessment document.

Edited March 31, 2017 by Curly