Fixing Main Gun Zeroing Parameters

[harrypanaich]

Its been long since I last posted something worthwhile, but now I need help in addressing grey areas in a paper I am writing on 'The Science and Art of Main Gun Zeroing'.

 

I know, and please correct me if I am wrong, that the primary criteria for fixing a range for zeroing the main gun of a contemporary MBT is as under:-

• The ammo used must be the primary tank-killing ammo of the MBT, usually FSAPDS.
• The size of the zeroing target must be fixed before hand.
• Based on the dispersion data for ammo used for zeroing, the hit probability must be at least 90%.

The M1A1/A2 tanks zero their main gun at 1500 meters. While I do not have the dispersion data for the 120mm ammo used on the M1A1/A2 for zeroing, I do have the zeroing ('screening') target dimensions as under :-

 

 

I presume that the dia of the 'bull' ( is based on a statistical assurance of at least 90% hits or that each round fired has a 90% chance of hitting within the circle.

 

This brings me to my primary reason for posting here. If I have the dispersion data of a particular ammo variant, say BM-9/BM-12 (at http://www.kotsch88.de/tafeln/st_125mm-ke.htm) can I work out the theoretical minimum zeroing range where at least 90% hit probability is assured? I have tried my hand at working this out but I am not sure if I am right, or if I am missing something.

 

The 'Range Table' (or Firing Table) at http://www.kotsch88.de/tafeln/st_125mm-ke.htm mentions the Mean Deviation figures (in meters) for dispersion for 125mm FSAPDS (BM-12) ammo seperately in elevation and azimuth, in 100 meter range increments. The English language translation of the header row of the table at the ibid web page is attached [attachment=1293:Range Table Header Row.png].

 

My understanding of the process to work out a zeroing range was as under:-

1. For each range increment in the Range Table, I convert the Mean Deviation to Probable Error (= 0.84535 x Mean Deviation). This conversion factor is applicable for Normally Distributed data.
2. The 50% Zone is twice the Probable Error (= 2 x 0.84535 x Mean Deviation). Again, this conversion is applicable only for Normally Distributed data.
3. The predetermined size of the zeroing target is fixed as 1.2 meters x 1.2 meters (4' x 4') for working out the corresponding 90% Hit Probability zeroing range.
4. So I work out the Probability Factor (= Target Dimension / 50% Zone)
5. I refer to the 'Table of Linear Probability Factors' [attachment=1296:Tble of Linear Prob Factrs.png] and read the corresponding hit probability against tabulated Probabilty Factor value computed above at Step 4.
6. I work out the Hit Probabilities seperately for the dispersion data in the elevation plane and in the horizontal (azimuth) plane.
7. The cummulative Hit Probability at that range is P_Hit = P_Elevation x P_Azimuth
8. I repeat the entire process (using Excel) for each 100 meter range increment. Refer to the attached Excel sheet [attachment=1294:BM12 Hit Prob.png]
9. Going by my calculation for a 1.2 meter square target, the Excel sheet shows 90% hit probability is assured at 900 meters firing BM-12 ammo from a D-81 tank gun.

Therefore, the prescribed zeroing range for D-81 tank gun firing BM-12 FSAPDS ammo is 900 meters.

 

Am I correct in my understanding of the statistical science and procedure for arriving at a zeroing range for a particular ammo-gun combo ? Please guide, comment and advise. Thank you in advance.

 

I have refered to the following books to try and learn the theory:-

• Textbook of Ballistics and Gunnery, Volume One (Her Majesty's Stationery Office (HMSO), London, 1987).
• Rheinmetall Handbook on Weaponry, Second English Edition 1982 (Rheinmetall GmbH).
• Oerlikon Pocket Book, Second Revised Edition 1981 (Oerlikon-Bührle AG, Zurich).
• Technology of Tanks, Volume One, 1991, Richard M Ogorkiewicz (Jane’s Information Group).
• Ballistics and Ammunition, Canadian Army Field Artillery Pamphlet, Volume 6, B-GL-306-006/FP-001, 1992.

[Mobius]

Looking at your table and asking: Isn't the std. deviation larger than the 50% zone? It covers 68.27% of the area under the curve not just 50%,

[harrypanaich]

I have used the traditional (as in field gunnery, actually) concept of 'Probable Error' and converted the same to a '50% Zone'. The standard deviation σ is mentioned in the table to aid other who may use the derived figure for scrutiny of my post.

 

In the context of Normal Distributions the inter-se conversion factors are cross-tabulated as under (courtesy the Rhienmettal handbook);

 

η σ γ

Mean Deviation, η 1 0.79788 1.18294

Standard Deviation, σ 1.25331 1 1.48260

Probable Error, γ 0.84535 0.67449 1

 

From the above σ = 1.25331 η or 50% Zone ρ = (2γ) = 1.3490 σ

 

The conventional statistical procedure is to form the standardised deviate and use the table for the standard normal distribution. Thus for a zone extending a distance d on either side of an MPI, the standardised deviate z = d / σ then the chance of the round falling in the zone is 2A(z) where A(z) is read from the table.

 

If consistency is specified in terms of the 50% zone then the required solution can be read directly from the Table of Linear Probabilities. In this Table, the chance of a round falling within a given zone is read as a percentage corresponding to the value of a probability factor f, where f = d / ρ (ratio of the zone length d being considered to the length of the 50% zone ρ)

Edited April 14, 2013 by harrypanaich

[Mobius]

Your target is 4' x 4'. To get a good zero the center should be at the gun's height.

[harrypanaich]

Your target is 4' x 4'. To get a good zero the center should be at the gun's height.

 

I agree ideally the trunnion height and the centre of the zeroing target should be dead level, on the same plane.

 

But your observation is a matter of idealised procedure. It's the theoretical science that I am trying to come to grips with at this stage.

 

I need answers to zeroing questions, for instance in the context of the M1A1/A2 - why is the 120mm gun zeroed at 1500 meters using a 1.75 meter circular 'bull' ? Why not at 2000 meters or 1900 meters ? Why not with a smaller or a larger 'bull'?

[harrypanaich]

The answer is out there somewhere ! Come on folks - this is 'bread and butter' stuff for alot of us.

 

I am also looking for the US Army Materiel Development And Readiness Command, Engineering Design Handbook. Army Weapon Systems Analysis. Part 1 (DARCOM-P 706-101). I got the Part 2 version, it has nothing on the hit probability aspect that I seek. The Part 1 is avaialable as a paid version and I am not too sure of its content.

[Mobius]

I don't think you are looking at this right. Look at it physically. The BM-9 50% zone at 1000m is .27m across. The 50% zone is 0.67449σ. 1 std deviation is 1σ. So std deviation is .4m across. 68% of the hits will be within .4m of the center of the target. Draw a circle or box this size around the midpoint of 1.22m square target. The entire target covers 1.22/0.4 std deviations. That is over 3σ so 99.7 % lie within the target. Ok, its 99.7 x 99.7.

 

http://en.wikipedia.org/wiki/Standard_deviation

 

That Rhienmeital handbook is showing the 50% twice as big as the wiki and other math sites. I wonder why.

 

[Edit] OK, now I see why. The zero number is the entire width of zero box not just the distance from the mean as in the wiki table.

Edited April 15, 2013 by Mobius

[Stefan Kotsch]

deleted ...

Edited April 15, 2013 by Stefan Kotsch

[harrypanaich]

I don't think you are looking at this right. Look at it physically. The BM-9 50% zone at 1000m is .27m across. The 50% zone is 0.67449σ. 1 std deviation is 1σ. So std deviation is .4m across. 68% of the hits will be within .4m of the center of the target. Draw a circle or box this size around the midpoint of 1.22m square target. The entire target covers 1.22/0.4 std deviations. That is over 3σ so 99.7 % lie within the target. Ok, its 99.7 x 99.7.

 

Actually if you check out the footnote to the cross-tab at Posr #3 the 50% Zone = (twice the Probable Error) = 1.3490 σ

 

Apropos my remark at #6 on the liklihood of the subject being elaborated upon in US Army Materiel Development And Readiness Command, Engineering Design Handbook. Army Weapon Systems Analysis. Part 1 (DARCOM-P 706-101), actually a lot of published references point to an authoritative paper 'On the Computation of Hit Probability' by Hermann Josef Helgert. I have the same author's paper on 'A Statistical Treatment of Various Classes of Gunner Errors & the Calculation of Hit Probability' at http://www.dtic.mil/dtic/tr/fulltext/u2/697697.pdf

 

There was another reference that I read somewhere (I am still trying to recall / retrieve the same) that mentioned that stipulating a zeroing range merely on statistical hit probability assurance of greater than 90% is an academic excercise and needs to take into account the lethality aspect (kill probability) when referring to a weapon system as a whole. It sounds rational but I am not convinced - I need authoritative references, explanations and procedures.

[DKTanker]

I need answers to zeroing questions, for instance in the context of the M1A1/A2 - why is the 120mm gun zeroed at 1500 meters using a 1.75 meter circular 'bull' ? Why not at 2000 meters or 1900 meters ? Why not with a smaller or a larger 'bull'?

Technically the M1/A1/A2 do not zero their main guns, they confirm system calibration through a screening process.

 

The reason for the 1.75m circle is to the gunner it will appear to be the same size as if it were the 1200m target (1200m target uses 1.4m circle).

As for why 1500meters. At least two fold. One is legacy, the previous screening process used targets at 900, 1200, and 1500 meters. The second reason is to ensure the M865 remains stable (M865 starts becoming unstable beyond 2000m). Other possible reasons include but not limited to range depth limitations and a longer range than 1500m proves nothing more than what happened at 1500m.

[Mobius]

The target has to be considered above and below the midpoint. Thus the 50% zone has to be a relationship to 2σ. So back to the 0.67449σ. IRQ = 0.67448975

Edited April 16, 2013 by Mobius

[harrypanaich]

Technically the M1/A1/A2 do not zero their main guns, they confirm system calibration through a screening process.

 

The reason for the 1.75m circle is to the gunner it will appear to be the same size as if it were the 1200m target (1200m target uses 1.4m circle).

As for why 1500meters. At least two fold. One is legacy, the previous screening process used targets at 900, 1200, and 1500 meters. The second reason is to ensure the M865 remains stable (M865 starts becoming unstable beyond 2000m). Other possible reasons include but not limited to range depth limitations and a longer range than 1500m proves nothing more than what happened at 1500m.

 

Thx for responding DKTanker but my query remains factually unanswered.

 

While I concede the aspect of the M1A1/A2s relying on calibration by firing at a 'screening' target, in essence it is akin to zeroing. The elaborate fire control system allows the tank crew to punch-in the offset or variations between the Point of Aim (or Expected Point of Impact) and the MPI.

 

The dimensions of the 'screening' target or the zeroing target are co-related to the range and thence to the 'hit probability' aspect. Forgive for differing on this aspect, but I don't think legacy issues are involved.

 

I am appending a page from the FM 17-12 (Sep 1978) on zeroing the M60A3 as an attached file ('M60 Zeroing.jpg'). Here too, apart from variations in the zeroing procedure, the end-state refers to achieving the MPI within a 24" dia circle on a target placed at 1200 meters.

 

 

I have yet another example that I worked on in trying to see if the theory of zeroing fits all sizes. At the face of it, it does but I guess it's a fluke. I need informed comments and references on the why's that bedevil me.

 

Zeroing Target for 105mm Tank Gun The 105mm APDS ammunition fired by the L7 gun had a dispersion of 0.18 mil standard deviation. The prescribed zeroing range for the 105mm tank gun (on British MBTs) was 1100 meters.

We convert the angular measure (mils) of dispersion into a linear measure (meters) standard deviation, σ, at 1100m as under :-

0.18 mils = 1000 (σ / 1100) or σ = 0.18 x 1.1 = 0.20 m

Going by the dictum of P_H=90% at the zeroing range, P_H can be split up into its elevation and azimuth components P_E and P_A as under, assuming P_E = P_A :-

P_H = P_E x P_A or 0.90 = P² or P_E = P_A = 94.9%

From the Table of Linear Probabilities, for a hit probability of 94.9% the probability factor, f = 2.91.

As Probability factor, f = Target Height, d / 50% Zone

f = 2.91 = d / ρ = d / (1.3490 σ) [as 50% Zone = 1.3490σ]

2.91 = d / (1.3490 x 0.20) = d / 0.27

d = 0.27 x 2.91 = 0.79 m or 31.1” ≈ 36”

 

Since P_E = P_A was assumed, the target width and height would be identical, thereby suggesting the use of a 3’ x 3’ zeroing target. QED

Edited April 16, 2013 by harrypanaich

[Mobius]

The solution works out on the standard distribution curve. Where σ is .2m from the mean.

Then 2σ of the the mean is 0.4m and would be 95.45%. So very close to 0.395m which would be 94.9%.

 

Another solution would be to find where 94.9% was under the distribution curve and then find the number x σ that this represents. That would save a few steps.

1. 90% = square of 0.94868.

2. Per z score (area under the curve from table of 0.94868) = (+/- 1.9488 σ)

3. 0.2m x 1.9488 = .38878m.

4. Since area below and above mean must be added together = 2 x 0.38878m =0.77952m or 30.6".

Edited April 16, 2013 by Mobius

[DKTanker]

harrypanaich, on 16 Apr 2013 - 03:23, said:

Thx for responding DKTanker but my query remains factually unanswered.

To that end, might I advise you to no longer conflate tank gunnery doctrine for an M60A3 in 1978 with that of an M60A3 in 1981-present, or an M1 of any type at any time?

 

Next, there is no science or art of zeroing the M60A3 or the M1/M1A1/A2. It isn't done and hasn't been done for thirty years*, and no, screening and zeroing are not akin to each other. One attempts to discreetly change the point of aim for a particular type of ammunition, of a particular lot, fired at a particular target at a particular time under particular environmental condidtions utililzing multiple rounds of ammunition. The other merely confirms, with one round, that the fire control system, ammunition, and gun with mount are all operating correctly and in concert. If you want to speak of the art and science of main gun zeroing, you'll have to find different tanks which use different doctrine.

 

*There is the art of bullshit for those tanks, I know, I did it. But everytime somebody said such and such tank had to have a discreet zero, it was found that the there was a FCS problem, or a mechanical problem, or that there was a problem with a particular lot of ammunition.

[Colin]

I suspect that each tank would have it's own quirks, with different amounts of slop in the mechanical components that the crews would become aware of.

[DKTanker]

I suspect that each tank would have it's own quirks, with different amounts of slop in the mechanical components that the crews would become aware of.

You would think so, not really the case. There are so many variables involved, each with their own tolerance to absolute, not to mention human error, for each individual round fired, that trying to establish a discreet zero for each type and lot of round has become a fools errand. Putting aside gross mechanical, electrical, and ammunition problems there is still the system error budget with which to be concerned.

 

For instance, let us suppose that the tolerable mechanical budget error for gun jump is +/- .1 mil, that the system elevation error is +/- .1 mil, that the azimuth error is +/- .1 mil, that the superelevation input is +/- .1 mil....see where I'm going? Then there is the muzzle jump error due to interior ballistics of the gun...suppose it is also +/- .1mil, and then there is the round to round dispersion tolerance of about +/- .2 mil. Of course there are environmental factors involved such as crosswind, ammo temperature, air temperature, barametric pressure, direction of firing in relation to earth, the aforementioned target height relative to gun trunnions (is that necessary, or should it be relative to the muzzle with the appropriate SE applied?). Each of those elements has its own tolerance to absolute, and we haven't even discussed the human error budget, which some argue is as much as .3 mil. If you add up all those tolerances you can quickly get to 1 mil of angle or greater error...and that is for just one round fired. Sure, most of those errors will randomly tend to , if not cancel out, then reduce the amount of total error, but you can never be sure, you can never know how much, and you can never know in what direction.

 

It is with that ratonale that the US Army ceased zeroing tank main guns of M60A3s and M1s over thirty years ago.

[harrypanaich]

 

harrypanaich, on 16 Apr 2013 - 03:23, said:

Thx for responding DKTanker but my query remains factually unanswered.

To that end, might I advise you to no longer conflate tank gunnery doctrine for an M60A3 in 1978 with that of an M60A3 in 1981-present, or an M1 of any type at any time?

 

Next, there is no science or art of zeroing the M60A3 or the M1/M1A1/A2. It isn't done and hasn't been done for thirty years*, and no, screening and zeroing are not akin to each other. One attempts to discreetly change the point of aim for a particular type of ammunition, of a particular lot, fired at a particular target at a particular time under particular environmental condidtions utililzing multiple rounds of ammunition. The other merely confirms, with one round, that the fire control system, ammunition, and gun with mount are all operating correctly and in concert. If you want to speak of the art and science of main gun zeroing, you'll have to find different tanks which use different doctrine.

 

*There is the art of bullshit for those tanks, I know, I did it. But everytime somebody said such and such tank had to have a discreet zero, it was found that the there was a FCS problem, or a mechanical problem, or that there was a problem with a particular lot of ammunition.

 

By referring to the 'science' of zeroing I was merely highlighting the aspects of selecting a particular range for 'zeroing', fixing the target size (or the zeroing panel) and the 'acceptance criterion' which decides that one tank IS zeroed whereas another IS NOT. These specifics of range, target size etc are based on a technical consideration of the trajectory, dispersion, lethality (presumably) characteristics of a particular gun-ammo combination. By attributing a scientific facet I implied its amenability to quantification.

 

The 'art' of zeroing referred to the procedural aspects tank crews are supposed to adhere to strictly in terms of ensuring equipment preparation, range drills, boresighting accuracy and crew competence. These aspects cannot be quantified and hence their generalisation as an 'art'. Apropos your reference to a 'discreet zero', despite its relative success on occasions, plays havoc with the entire effort of collective field firing in throwing up irrational firing results that go against the overall firing performance of the particular ammo lot, tank fleet or past firing record. Anyways these are semantics and aren't the moot point here.

 

Somewhat similar to the Americal practice of 'screening', some armies' tank fleets follow the practice of 'check-zero' which is a test of the retention of zero by a tank that has not fired for some time. I may be entirely wrong in my understanding, but I gather that the practice of 'zeroing' in the US parlance was affected by range toxicity concerns arising after adoption of DU ammunition, the costs of zeroing each tank afresh each time it fired, as well as the reliability and sophistication of contemporary FCS.

 

My research is aimed at the T-72 family of MBTs which have a fairly primitive FCS (by today's standards) despite an integral LRF.

 

The questions remain...

 

Whereas a standard NATO head-on target is 2.3 meters square, why is the zeroing / screening target panel disproportionately smaller ?

 

Whereas the effective range (and the maximum point-blank range, MPBR) of a contemporary 120 mm or 125 mm FSAPDS round is approximately 2100-2300 metres (generalised for a MV above 1600 m/s), why is the zeroing range, or that for screening, restricted to substantially lower ranges ?

 

The reasons for the above can't be legacy, surely ? That is what I have no idea about and seek your wisdom.

 

 

[Mobius]

So it seems for 1000m your 125mm zeroing get the std. dev. from the mean dev, or 0.27m x .79799 = 0.2154

Then plugging into #3 and #4 my equations gets:

3. 0.2154m x 1.9488 = .41978m.

4. Since area below and above mean must be added together = 2 x 0.41978m =0.818m or 32".

[harrypanaich]

Actually I have found considerable coverage of what we have been speaking of in a paper at http://www.dtic.mil/dtic/tr/fulltext/u2/a193618.pdf (Predicted Effect of Projectile Dispersion on Target Hit Probabilities & Dispersion Zone Sizes for the 25mm Gun of the Bradley Fighting Vehicle'. The applicability of the theory of hit probability as pertaining to a cannon burst fired by an IFV vis.a.vis single round fired by a tank gun has to be rationalised I guess.

 

Let me give it a go and see if I can see light! Meanwhile I welcome additional analysis and comments from all those who have graciously and generously shared their wisdom here.

[Mobius]

Actually I have found considerable coverage of what we have been speaking of in a paper at http://www.dtic.mil/dtic/tr/fulltext/u2/a193618.pdf (Predicted Effect of Projectile Dispersion on Target Hit Probabilities & Dispersion Zone Sizes for the 25mm Gun of the Bradley Fighting Vehicle'. The applicability of the theory of hit probability as pertaining to a cannon burst fired by an IFV vis.a.vis single round fired by a tank gun has to be rationalised I guess.

Thanks for the link. It gave me something to check against my program results. Spot checking I get pretty close. within at most 1.4%. But most of the time with rounding within 1 %. I did find what might be a typo on Table 1, 6x6 target 1000m, .3 dispersion (97) is less than .35 dispersion (98). Or, it may be that 1983ish IBM program is using a standard deviation algorithm. I know I gave up on the one I was using and now brute force it with a table lookup.

I have mostly dealt with dispersions of WWII guns, and those seem to have more dispersion in the vertical than the horizontal. All these modern guns seem to have equal dispersion in either. I wonder if that is because they are fin stabilized rather than spin stabilized?

[Old ROF]

Thought these might be of interest about fleet zero methods if not already known.

 

http://www.armyconference.org/ACAS00-02/ACAS00/WebbDavid/WebbDavid.pdf

 

http://ciar.org/ttk/mbt/armor/armor-magazine/armor-mag.1995.mj/3zero95.pdf

[arrow]

Zeroing is a tank gun calibration policy which was used with tanks without fire control computer. Zeroing was made on the same distance as boresighting. The standard range was 1000 yards up to 1500 meters. A big drawback of this policy is that a once established zeroing don’t remain constantly in the course of the time. I posted in “Long Range Tank Gunnery – Again" (post #103) an example with zeroing variations within a couple of weeks. Since the late 80ies we are using in Switzerland successful the fleed zero. The zeroing correction has to be fixed only with a few tanks and is o.k. for the whole fleet.

 

Back to individual zeroing.

The distance should be according an average combat range which has to be the boresighting distance too. This distance has nothing to do with dispersion. The zeroing target has to have appropriate dimensions.

 

Hit Probability

The round to round dispersion can be well described with a normal distribution. With it we are able to calculate hit probabilities. Because the calculation is not so easy a lot of tables exist. But often it is very cumbersome to get the desired probabilities. Excel provides some powerful statistic functions which make the calculation very easy. Attached is an Excel sheet which contains 2 tools to determine hit probabilities or target size for a desired hit probability:

 

Hit Probability for Rectangle Targets

Calculation of the hit probability for rounds fired on rectangle targets. Separate target size and dispersion in azimuth and elevation are possible. The aiming point may vary from the mean point of impact.

 

Hit Probability immediately after Zeroing

Calculation of the hit probability for the check round after zeroing with a group of any number of rounds at a rectangle target. For a requested hit probability the related target radius is calculated.

[Guest crockett007]

Ok so speaking from the perspective of a tank gunner, I can't really find a practical use for this information. POI is affected by too many variables in field application. Bore sight and zero is performed under stationary and controlled conditions. Mostly a "get the round on target exercise". For the Abrams, the original zero is moot after the first application of the MRS. Also, I can tell you from direct experience, using sabot at long range in the desert, that a 1200-1500m zero POI, was a different animal when engaging hull down targets at 3000m and beyond. I don't claim to know the iterations in the ballistic solution software that contribute to this.

 

I have re adjusted my zero in combat at extreme range on destroyed vehicles to compensate for this fact. Also, shooting on the move is another variable that needs to be considered. Zeroing in Normal mode in a stationary posture is not relevant to shooting on the move under stabilization. We constantly adjusted our FC in the field to compensate for the conditions at hand.

 

The problem with statistical probability is that the calculations are theoretical and not practical. Shooting tanks in combat is about as practical as one can get.

Edited May 21, 2013 by crockett007

[harrypanaich]

Zeroing is a tank gun calibration policy which was used with tanks without fire control computer. Zeroing was made on the same distance as boresighting.

 

Back to individual zeroing.

The distance should be according an average combat range which has to be the boresighting distance too. This distance has nothing to do with dispersion. The zeroing target has to have appropriate dimensions.

 

 

 

The problem with statistical probability is that the calculations are theoretical and not practical. Shooting tanks in combat is about as practical as one can get.

 

I seem to have probable mixed up the issues. Apologies.

 

Firstly ballistic zeroing is applicable to tanks that lack a full-solution FCS. Tanks like the T-55, T-72 and the Chinese equivalents need to be zeroed 'periodically'. Of that there is no doubt !

 

I need answers to explain the fact that why is zeroing (or calibration) carried out at a fixed distance that is usually less than the maximum effective range of the ammunition used for zeroing ?

 

Secondly, what is the association between the dimensions of a zeroing target (or screening target) and the range at which it is placed from the tank ? For instance in the M1 series the screening target is placed at 1500m and the 'bull' circle measures 175cm (refer to the diagram on my first post here). Why ? Is it a statistically derived target dimension based on hit probability , or is it some other factor ?

 

Thanks.

[Lieste]

It seems to me that there are four basic tasks being performed here.

 

Zeroing ~ deciding where to minimise mean error from the aim point (that error usually small at most ranges - small errors in MPI have a largest impact at intermediate ranges where the pH field is sufficiently 'peaky' that the centre location has some discernable effect and not so 'peaky' that only gross errors will force a miss)

Boresighting ~ making sure that the gun aimpoint is shifted appropriately with inputs from the FCS and gunner's controls - ensures that the FCS returns predicted behaviour under standard conditions.

Updating MRS ~ aligning the bore's muzzle and sight so that the starting direction for shells isn't grossly affected by temperature of the tube

Screening - firing a 'zeroing' round that confirms/questions the success of the above steps. There will be false positives, false negatives and 'good' values... the trick is to select a distance and size that doesn't 'fail' too many good setups, but catches as many poor ones as possible.

Even when done correctly there will be some that shoot well, and others badly.