Jump to content


Photo

Definition "probable Error" In U.s. Firing Table, Ca. 1960


  • Please log in to reply
18 replies to this topic

#1 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 06 August 2019 - 1841 PM

I have a firing table of US origin, ca. 1960, that gives me a "probable error" in range and deflection. How is that defined, as the 50% quartile, or as one, two, three standarddeviations? Or am I completely off base?


  • 0

#2 DB

DB

    Crew

  • Members
  • PipPip
  • 11,232 posts
  • Gender:Male
  • Location:Hertfordshire, England

Posted 08 August 2019 - 1417 PM

"Circular Error, Probable" would be a similar metric and is the 50% percentile, so it seems likely to mean 50% here, too.

 

Otherwise, they'd be using Standard Deviations with 1SD for a normal distribution being ~68%.


  • 0

#3 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 08 August 2019 - 1638 PM

The problem I'm having with that explanation - which sounds plausible at first sight - is that this would actually mean a HUGE ammo spread (in the order of magnitude of "2 mil" for one standarddeviation, which appears hard to believe).

 

If it were 3SD, this would be much more in line with conventional ammo. Admittedly this is for a recoilless rifle and for the typical engagement ranges this might not matter all that much. But then the question is, is it "the ammo" that turns this contraption into a scatter gun, or is is the whole gun assembly and limitations of the sight. For a firing table I would assume that it just accounts for the composite error of the gun assembly and the ammunition itself, assuming an otherwise perfect aim.

Admittedly recoilless rifles are somewhat the "hobgoblin attempt" at precision fire so I'm not expecting a .25 mil standarddeviation. But it's difficult to reconcile anything in between with a reasonable mathematical definition. The 50% quartile implies a giant spread, 3SD (=95% of all shots) implies a spread comparable to high performance tank guns. Both approaches are not entirely satisfactory.


  • 0

#4 DB

DB

    Crew

  • Members
  • PipPip
  • 11,232 posts
  • Gender:Male
  • Location:Hertfordshire, England

Posted 09 August 2019 - 2120 PM

3SD would be 99%+
2SD is 95%

I can't answer your question, I'm afraid. 2 mil is 7 inches or so at 100 yards. Is that really "huge"?

Infantry rifles were good enough if they were 2 moa or so, in wartime at least, which is only 3.5 times better.
  • 0

#5 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 10 August 2019 - 0119 AM

When the engagement ranges are given in "several kilometers", 2 mil spread for a 50% chance to miss despite perfect aim should be "huge" by anyone's definition, I'd go so far to call it "hugeless".


  • 0

#6 bojan

bojan

    Crew

  • Members
  • PipPip
  • 11,542 posts
  • Gender:Male
  • Location:Belgrade, Serbia
  • Interests:Obscure tanks and guns.
    Obscure facts about well known tanks and guns.
    Obscure historical facts.

Posted 10 August 2019 - 0255 AM

I can't answer your question, I'm afraid. 2 mil is 7 inches or so at 100 yards. Is that really "huge"?
 

For a large guns - yes:

 

100mm T-12:

- APFSDS - .25 mil

- HEAT - .21-0.25.

 

US 90mm M3:

- T33 AP - 0.16 vertical, 0.12 horizontal 

- M71 HE - 0.17 vertical, 0.13 horizontal

 

Compare to that Yugo M60 82mm recoilless:

- M60 HEAT - 0.70

- M70 HEAT - 0.76


  • 0

#7 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,746 posts
  • Gender:Male
  • Location:Where we don't bow to the Queen
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 10 August 2019 - 1536 PM

I'm pretty sure US probable error is 50%.


Edited by Mobius, 10 August 2019 - 1540 PM.

  • 0

#8 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 11 August 2019 - 0254 AM

...but it's the diameter of a square region, not a radius (or, technically, two half-axes of an ellipse), right?


  • 0

#9 Interlinked

Interlinked

    Crew

  • Members
  • PipPip
  • 289 posts
  • Gender:Male
  • Interests:Cold war era tanks

Posted 11 August 2019 - 0920 AM

It should be the square region where 50% of the shots land. Same definition as the Soviets, no?


  • 0

#10 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,746 posts
  • Gender:Male
  • Location:Where we don't bow to the Queen
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 11 August 2019 - 1627 PM

For the Soviets it is the radius. It is not square like the Germans.

https://helpiks.org/4-1512.html

 

I think the US uses the radius as well.  In post war US documents I've seen it in standard deviations and in mils.  

https://apps.dtic.mi...t/u2/316221.pdf
https://apps.dtic.mi...t/u2/065653.pdf

 

I've seen the British use 50% radius and 90% in WWII.


Edited by Mobius, 11 August 2019 - 1628 PM.

  • 0

#11 FALightFighter

FALightFighter

    Red-Legged Ilk

  • Members
  • PipPip
  • 1,416 posts
  • Gender:Male
  • Location:USA

Posted 11 August 2019 - 2128 PM

Probable Error is a 1 standard deviation in artillery firing tables - I assume that others are the same.
  • 0

#12 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 12 August 2019 - 0259 AM

Yay. More and more options.

 

But what was used specifically by the US Army, specifically for direct fire weapons, specifically around 1960?


  • 0

#13 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,746 posts
  • Gender:Male
  • Location:Where we don't bow to the Queen
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 12 August 2019 - 0856 AM

Yay. More and more options.

 

But what was used specifically by the US Army, specifically for direct fire weapons, specifically around 1960?

The US reports I've seen use the word dispersion for shell deviation from 1945 to 1990. 

But from that first pdf I listed.

 

 

Dispersions are given in terms of their standard deviation (a), rather than probable error.

 

Before this the US went with the British mean point of impact (MPI).

 

Ssnake, what gun is this?  Maybe I have come across it.


Edited by Mobius, 12 August 2019 - 1401 PM.

  • 0

#14 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 12 August 2019 - 1721 PM

106mm Recoilless Rifle (as mentioned), M346 HEP-T.

 

e.g. for 1000m range the probable error in deflection is given as 1.0 mil

This remains pretty much constant out to 4km range, after that it increases a bit.


Edited by Ssnake, 12 August 2019 - 1722 PM.

  • 0

#15 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,746 posts
  • Gender:Male
  • Location:Where we don't bow to the Queen
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 13 August 2019 - 0910 AM

106mm Recoilless Rifle (as mentioned), M346 HEP-T.

 

e.g. for 1000m range the probable error in deflection is given as 1.0 mil

This remains pretty much constant out to 4km range, after that it increases a bit.

It would seem that it would put it in the MPI and CEP camp of a fraction of the range.

In another report I found the following definition.  It looks like the area would be the 50% area but the radius would the standard deviation for that.

 

.. common parameter for. describing the accuracy of a weapon is the circular probable error, generally referred to as CEP. CEP is simply the bivariate analog of the univariate probable error and measures the radius of a mean-centered circle which includes 50% of the bivariate probability. In the case of circular normal errors where the error variances are the same in both directions, CEP can be expressed as a function of the common miss distance standard deviation.

​


  • 0

#16 DB

DB

    Crew

  • Members
  • PipPip
  • 11,232 posts
  • Gender:Male
  • Location:Hertfordshire, England

Posted 13 August 2019 - 0944 AM

Not sure that's what the quoted reference says.

Given a normally distributed measure, then the Probable Error is indeed a *function* of the SD, but it's not the *same* as the SD, which is what you seem to be saying.
  • 0

#17 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,746 posts
  • Gender:Male
  • Location:Where we don't bow to the Queen
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 13 August 2019 - 1034 AM

Not sure that's what the quoted reference says.

Given a normally distributed measure, then the Probable Error is indeed a *function* of the SD, but it's not the *same* as the SD, which is what you seem to be saying.

I agree that the wording is confusing.  But, average mean and probable error are all functions of standard deviation.  But, why mention it?   The WWII German method of 50% zone is 1.349 SD.

img37.gif


Edited by Mobius, 13 August 2019 - 1039 AM.

  • 0

#18 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,587 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 14 August 2019 - 0112 AM

Thanks, Mobius.


  • 0

#19 Stefan Kotsch

Stefan Kotsch

    Crew

  • Members
  • PipPip
  • 849 posts
  • Gender:Male
  • Location:Ludwigslust, Germany
  • Interests:Tanks - fire control, guns, ammo, autoloader

Posted 17 August 2019 - 0745 AM

deleted


Edited by Stefan Kotsch, 17 August 2019 - 0757 AM.

  • 0




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users