Schneider 1917 155mm howitzer

[LCo308]

Trying to get an idea of the typical accuracy of this gun in WW1. Maybe a general CEP as a function of range, or something similar.

Also wondering what additional errors resulted from the absence of an OP with unobstructed view of target and resulting need to fire off of limited range fire corrections against a separate known target with unobstructed LOS to the OP.

[Spaceman]

By the time the C17S entered service, it was mainly used for predicted fire - this technique relied on both the gun's position and the target position being precisely known on a map through surveying and aerial photography so that the gun could be aimed at the target without the need for forward observers or ranging shots.

This weapon was normally used to attack enemy trenches and lines of communication. Since predicted firing was mainly employed, there was a need for these howitzers to provide a reasonably sized beaten zone covering the target area rather than trying to land consecutive shells as close together as possible. In hitting these types of targets, what was important was to take into account the meteorology conditions, the shell weight, the barrel wear and the muzzle velocity that all greatly affected the accuracy of the weapon (and therefore where the beaten zone lay) as opposed to the intrinsic precision of the weapon (to which a CEP relates) that determined the size of the beaten zone.

 

[LCo308]

Thanks for the quick response Spaceman. I appreciate your insight on this topic.  Your comment regarding the need for a reasonably sized beaten zone is interesting. Was the desired spread in fall of rounds the result of inherent variabilities of the weapon, fixed adjustments for the mission amongst the different guns in a battery or were adjustments "dialed in" during a fire mission by the crew of each gun?  Also, what was the typical size/extent of the beaten zone, as a function of range (for both an individual gun or a battery of 3-4 guns).  I'm investigating the legitimacy of at least one (possibly two) friendly-fire/short-fire reports associated with a WW-1 fire mission.  Following are some particulars about the mission. I have also attached a sketch.  Looking forward to your feedback. -Jim

A nighttime (shortly after dusk) fire mission by two Schneider 155 batteries against two separate fixed targets with well-defined map coordinates (roughly 500 m apart) did not have an unobstructed OP for daylight registration against the targets.  Instead, just prior to sunset, observed registration firing (2 rounds per tube) was accomplished against a widely separate target ("T1" in the attached sketch) at a similar range, but with an approx. shift in azimuth of 25 degrees from the mission targets.  Errors in the fall of the rounds  fired against "T1" were noted from an OP roughly 5 km from the target.  No adjustments to fire, and therefore observations of corrected fire, were made.  I take it, this whole exercise was done to estimate "range" and/or "other" adjustments required for the existing meteorological conditions. (Don't know what the Met. conditions were or how they might have changed after sunset.)   Note: Leading up to and during this activity, enemy artillery was hitting the locations of the batteries and OP with significant amounts of HE and gas, making for a challenging observation environment.

Immediately after the above fire, the batteries were laid on the two mission targets, "T2" and "T3" in the sketch, I assume utilizing aiming stakes. (I suspect 3 guns per battery were used in the mission, with each battery addressing a separate target.)   The observer data from the registration fire against "T1" arrived at the batteries just prior to the deadline for the fire mission, which went off on-schedule.  Unobserved firing occurred over a 1-hr period, with 240 rounds launched against each mission target.  Note: one of the two targets (T2) was a machine gun position;  240 155mm rounds fired to take out one or two machine guns, amazing!

 

[ianjonesncl]

It sounds like a Witness Point Mission.

Fire on a known target in order to check the accuracy of fire without forfeiting surprise.

Then fire on the actual intend target where the effect is required.

Edited 13 July , 2023 by ianjonesncl

[LCo308]

I suppose it was equivalent to a Witness Point Mission, but surprise wasn't the driving factor here.  It was inter-visibilty problems between the existing OPs and the target(s). The mission targets were in a valley and no forward observers could view them due to intervening high ground.

I appreciate the commentary, but can you offer any citations for size of the "beaten zone" (per Spaceman) for a battery of Schneider 155's, especially under the circumstances I described?

 

[ianjonesncl]

1 minute ago, LCo308 said:

It was inter-visibilty problems between the existing OPs and the target(s). The mission targets were in a valley and no forward observers could view them due to intervening high ground.

Witness Point can also be used to determine the accuracy of metrological data - point you raised in previous posts.

If the OP observes the fall of shot on T1 to ascertain the adjustment to the firing data which could then be applied in the computation of the firing data for the unobserved targets T2 / T3. 

 

 

[Spaceman]

The spread of the fall of artillery rounds is referred to as dispersion that then defines the beaten zone on the ground. For an individual C17S  firing at a range of ~8 km, the dispersion was probably of the order of less than 50 m and was due to a range of variables including: errors in sighting between rounds, variability of the wind and other weather conditions, variability of the weight of the shells and propellants, and stability of the gun carriage during firing.

Dispersion effects are effectively random in nature. However, on top of this was the calibrated accuracy of the gun that resulted in the centre of the beaten zone not being centred at the target location. The largest source of this systematic error was the error in the knowledge of the gun's muzzle velocity that reduced over time because of wear in the bore. By 1917, calibration techniques had been devised to allow a gun's muzzle velocity to be determined which then made possible so-called predicted fire. In other words, there was no longer a need for target registration and ranging shots that were both expensive in ammunition usage and also removed the element of surprise - targets could be engaged purely from target coordinates derived from aerial maps.

The calibration technique was not perfectly accurate but, when considering battery fire, the individual errors just added to more dispersion. So if the dispersion from an individual gun was ~50 m, the dispersion from a battery might have been ~150m. 

For the target mission being described, I assume this involved the AEF and was therefore at a time when predicted fire had become the norm. It may well be that firing at target T1 was a calibration exercise for the guns. If the main mission was to destroy a pair of localised targets, the wasteful way of doing this would have been to fire off an excessive number of rounds that seems to be the case here. A less wasteful way would have been to use aerial reconnaissance to observe the effects of the shoot.

[LCo308]

I suppose it was equivalent to a Witness Point Mission, but surprise wasn't the driving factor here.  It was inter-visibilty problems between the existing OPs and the target(s). The mission targets were in a valley and no forward observers could view them due to intervening high ground.

I appreciate the commentary, but can you offer any citations for size of the "beaten zone" (per Spaceman) for a battery of Schneider 155's, especially under the circumstances I described?

 

[LCo308]

Thanks Spaceman; very informative.  This gives me a good sense for what the dispersion would be around an actual baseline impact point. 

Seems to me though that there would be additional errors associated with the establishment of that baseline point.  As you indicate, the fire against Target T1 is clearly a daytime calibration exercise for a nighttime firing (and not their first choice of registering against the actual targets, due to a lack of a suitable OP).  The accuracy of the registration fire against T1 is not reported, so we don't know the corrections required for proper calibration.  Due to disruption of telephone comms by enemy fire, we know that the OP involved received word of the registration firing by messenger just minutes before that firing began.  The messenger was an artillery officer, who reportedly arrived "with choked lungs and swimming eyes" from gas firings.  I assume he did not do the actual observing, but it seems conditions were hardly optimal for monitoring rounds falling roughly 5 km away.  Who knows what error this calibration effort added to the establishment of the baseline impact points for the two actual targets.

I also wonder about uncertainties in the establishment of azimuth, and less-so range, for the actual targets, even with the use of the accurately prepared maps of the time,  the targets not having been previously fired upon and not have a direct line-of-sight to either the guns or forward observers.  If the gunners are using alternative aim points for laying, even if the points go out, say 1 km, prominent features in this case are lacking, and I expect any extracted from a map, could be easily tens of meters "off", leading to ~100-200 m errors at say 8 km.  Perhaps such errors can be drastically reduced or compensated for through the use of a Director? 

Otherwise, what in your opinion, could be the source of a wild firing error (hundreds of meters) aside from a typographical error in a target coordinate within a written target order?  Thanks.

[ianjonesncl]

On 15/07/2023 at 18:00, LCo308 said:

daytime calibration exercise for a nighttime firing

I would suggest Calibration is a separate exercise which was undertaken by withdrawing the guns to a Calibration Range in order for the muzzle velocities to be measured. 

Registration is a methodology to determine the correction of the moment for variable factors to be applied to the firing data to determine the bearing and elevation ordered to the gun. 

 

 

Edited 6 August , 2023 by ianjonesncl

[Spaceman]

LCo308: I think the guns would have been calibrated in situ especially if this was being done to ensure a hit on a pair of targets at a particular range - it just requires a few observed rounds to be fired at a target at known range (T1) and, from that, to then determine the actual muzzle velocity from the observed error in the fall of shot. In general, the muzzle velocity would only change slowly over time due to the rifled bore wearing and therefore increasing the windage.

One of the obvious issues with this night-time firing exercise is the difficulty in laying for line. This would have been done using a dial sight in British parlance or panoramic telescope in US parlance or goniometre panoramique in French parlance. Essentially, when the guns were surveyed in, they would have been aligned with a Zero Line (expected mean direction to targets) and then the dial sight azimuth (slipping) scale set to zero with the sight centred on a suitable aiming point (AP) - this could be a prominent feature on the landscape or specially set up aiming posts. Target directions were then specified relative to the Zero Line direction and once set on the dial sight, it was only necessary to traverse the gun until the AP was again centred in the sight. In poor lighting conditions or at night, laying for line was not that easy. First off, the dial sight graticule had to be illuminated via a lamp to allow it to be seen in the sight. In general, daytime APs could not be seen and so it would have been necessary to use an aiming light instead. 

Laying for elevation would not have been difficult to do accurately at night but laying for line would have been more challenging. In the case of this particular fire mission, it would be interesting to know what actually happened during the firing and what went wrong. For example, was it a case of all the rounds missing the targets completely or was it just a few rounds that went astray from one particular gun? 

[ianjonesncl]

21 hours ago, LCo308 said:

source of a wild firing error (hundreds of meters)

One possibility it may be worth thinking about is a change in the metrological conditions. A memorable occasion I have witnessed  where fire was wide of the target occurred when the target was adjusted in wind and rain and fired in bright sunshine a about 3 to 4 hours later. I hasten to add it was not me who fired it.

One of the challenges with night firing is the change in air temperature as well as the other changes in metrological conditions. As an indication the British Artillery from March 1917 issued six telegrams daily, so once every four hours, for use in calculating predicted firing data.  The telegram issued at 18:00 was to be used from 19:00 to 23:00. It would seem sensible to be able to observe the effect of any new corrections for met prior to last light. 

If metrological data is not available, then without firing a co-ordinated illumination mission to adjust a witness / registration point, the last opportunity to have an indication of a metrological adjustment is before last light. If the metrological conditions change and you do not have updated metrological data, then you have to fire with what you have. No problem when you can adjust onto a target, if it is predicted data, then it is open to the errors you outlined. 

They are a myriad of reasons and some of them could be compounding, all the various factors @Spaceman has outlined. One error plus another error produces a bigger error.  As an estimate using the Gunnery Law of Sub-tension .....to be 200 metres adrift (excluding zone) at a range of 8,000 metres would require an 'error' at the Gun end of about 2 1/4 degrees. 

I would suggest night firing using predicted data on unobserved fire is at best harassing fire. It is literally a 'shot in the dark'. 

[LCo308]

Spaceman & Ianjonesncl:  Many thanks. Your comments have been very informative for someone with little knowledge of artillery basics.

Regarding: "what actually happened during the firing and what went wrong. For example, was it a case of all the rounds missing the targets completely or was it just a few rounds that went astray from one particular gun?"

It's known that Target T3 was completely destroyed, so at least several 155 rounds were on-target here, maybe more.

Some other "metrics" associated with the fire. 

1. It's known that two batteries were involved in the bombardment of Targets T2 & T3, with equal number of rounds fired against each target (240).  I figured it didn't make sense for all the guns involved to target both T2 & T3, but instead for each battery to fire upon a single target. 

2. The number of registration rounds fired from each battery against Target T1 was six.  This seemed an odd number of firings for four guns in a four-gun battery, so I figured  maybe three guns (2 registration rounds per gun) were being tasked from each battery, or perhaps two guns (3 reg. rounds per gun).  (A single gun with 6 reg. rounds just didn't make sense to me.)  During a one-hour period, 240 rounds were fired against Target T2 and 240 against Target T3.  If only 2 guns in each battery were involved, that would mean 120 rounds fired by each gun, meaning 2 rounds per minute for 60 mins.  If 3 guns were involved, 80 rounds per gun, or ~1.3 rounds per min.  I guessed that is was more likely the latter, resulting in 3 guns firing against each target.  Does that seem right to you?

3. If 3 guns actually targeted T3 and only one was on-target, its 80 rounds would have been more than enough to flatten it.  I suppose also, just one out of the three guns being wildly off (i.e., 80 rounds) would have explained the terrifying short-fire experience attributed by a doughboy to this bombardment.

 

 

 

[Spaceman]

Just to clarify the terminology: I believe we are talking about a single battery of 6 guns with each 3 gun section firing at a different target?

Again, in looking at the firing against target T1, this was probably a calibration mission rather than a registration mission. It might well have involved each of the 6 guns firing a single round to check the calibration accuracy of the gun and that the meteorological parameters had been correctly accounted for. 

Assuming the calibration of each gun was verified, it sounds as though each gun would have fired about 80 rounds. The C17S fired bagged cartridges and so the sustained rate of fire would probably have been just over 1 round per minute.

It would be interesting to know what the tactical disposition was of the troops involved on both sides were but I assume that, since the targets (T2 & T3) were machine guns, these guns may have been only a few hundred yards at most beyond the front line of AEF troops? If so, it would not have taken much of a sighting error in the dark of one of the guns for a round to fall perilously short?

 

 

[LCo308]

I am speaking of an American division here.  The artillery regiments were organized into battalions and batteries, more the equivalent of British batteries and sections (or "troops"?) I think.  There was a single artillery brigade in a division that consisted of two 75mm regiments and one 155mm regiment (24 guns each) with 3 battalions/regiment.  Each battalion (8 guns) consisted of two batteries (4 guns each). I will send you particulars on troop positions later. Have to run!  Thanks, Jim.

[LCo308]

Here are some graphics that hopefully provide a better picture of the problem (without going too far down the rabbit hole).  I've included a couple of troop positions; locations are approximate.  Of course there were soldiers all over these maps. I've only indicated those that are relevant to the story.

At the risk of sounding repetitive...  Given all the previously discussed challenges facing the observer at the OP shown in the first figure, I can't help but be skeptical about the adequacy of the "calibration fire" for accurate unobserved and uncorrected "predictive fire" of 480 rounds against the two mission targets an hour later after dark.    

[LCo308]

One additional question.

For what I now gather was likely "calibration" fire to adjust for the local weather conditions, 6 rounds were fired from each 4-gun battery. Would these 6 rounds come from a single gun (representing 3 short & 3 long)?  Was it common practice to determine proper adjustments for multiple guns using a single gun, assuming the nominal muzzle velocity of each gun was already known?  Thanks!

[ianjonesncl]

2 hours ago, LCo308 said:

One additional question.

For what I now gather was likely "calibration" fire to adjust for the local weather conditions, 

Calibration is not the process to adjust for local weather calculations. Calibration is the process to measure the muzzle velocity of the gun. The muzzle velocity at a point in time is constant, be it rain hail or shine.

This instruction issued in 1915 outlines what is required to Calibrate the guns. 

 

Calibration In the Field.pdf

[LCo308]

OK, sorry for using wrong terminology.  Assuming muzzle velocities are known, then firing must have been done to determine needed "adjustments" for existing met conditions.  Can you answer my query regarding the 6 rounds fired by each battery coming from a single gun?  Thanks!

 

[ianjonesncl]

4 minutes ago, LCo308 said:

OK, sorry for using wrong terminology.

Terminology is important as there are specific drills for different types of mission.

5 minutes ago, LCo308 said:

Can you answer my query regarding the 6 rounds fired by each battery coming from a single gun?

It would work with a single gun. The rounds would be adjusted to hit the target to determine the firing data. This would then be used to compare with the map data. The difference is the correction of the moment which includes meteorology.

Six rounds would seem to equate with what one expect from the adjusting one gun on to the target. At  most it could have been a section of 3 guns.... 3 rounds form one gun to adjust.... 3 guns one round. The advantage of using 3 gun is it allows a mean point of impact to be observed. giving the capability of a final adjustment

. 

 

[ianjonesncl]

These are the instructions regarding Registration.

 

 

Registration.pdf

[Spaceman]

The main inaccuracies in the gun fire would have been in the range direction and that incudes any errors in the calibration (MV) of the guns. Fortunately, the friendly forces seem to have been at least 200 m to either side of targets T2 and T3 and should therefore not have been in any real danger from the dispersion of the guns in that direction which should have been less than 50 m. The gunners would have measured the atmospheric temperature and pressure to take that into account and would have done the same for the wind direction. However, it would have made total sense to check the guns accuracy first against T1 with each gun firing a single round.

Although the stray rounds are described as short fire, that is not really the case because they are 0.5 km short but 1 km or 3 degs in deflection off target. It would be interesting to know how many stray rounds there were but I would suspect they were fired from the same gun which, for some reason, did not have his dial sight or panoramic telescope properly set up although this wouldn't explain why the rounds were also short? If this was the case, maybe this gun was not one of the ones that fired the 6 rounds at T1?

[LCo308]

Thank you both.

ianjonesnci, these were 4-gun American batteries. Are you saying in a 3-gun British section, it would have been normal practice (in such a situation) for 1 gun to fire 3 rounds and each gun to fire 1 additional round? If not, would typically have occurred?

Sorry for typo error. Meant to say "what would typically have occurred"?

[ianjonesncl]

The missions I used to conduct in modern times for Registration Missions to determine the correction of the moment, the normal drill was to conduct the mission with three guns from a six gun battery. You ranged onto a target with a single gun. Once you had fall of shot on the target , you then fired the three guns together to ascertain the main point of impact.  

That was the standard drill. The number of rounds fired varied.

Determining the correction of the moment could be done with one gun if required. I have done it, because the situation required it. They may had to do something completely different. 

What happened... happened. We can only guess what they were doing. 

It is a matter of applying gunnery principles to the tactical situation. 

 

 

 

[ianjonesncl]

Applying modern British artillery gunnery practices and any potential differences in drills between British and American methodologies to a situation that occurred over a century ago is challenging. The principles of operations are have however been the same for over a century. 

A thought of the events based upon interpreting the information in the original post and practical experience;

• Two Schneider 155 batteries were tasked to conduct fire during the hours of darkness against two separate fixed targets (T2 and T3) with well-defined map coordinates (roughly 500 m apart).
• They did not have line of sight from the OP to the targets for daylight registration against the targets. 
• The British Artillery circulated Meteor Telegrams which allowed corrections to be determined without recourse to firing. I do not know if the American Artillery adopted this method… but suspect they would have done.
• In order to ensure that the predicted fire data would be as accurate as possible, it would make sense that the correction to be applied to the map data from the battery to T2 and T3 to calculate the firing data be established or verified.
• The last opportunity to establish a new or verifying existing adjustment calculation would have been to view fall of shot before last light. This could only be done against a target that could be clearly identified on a map (T1)
• The methodology to verify that correction adjustment are;
  • Register Mission to calculate a correction
  • Witness Point mission to verity a correction
• When conducting fire missions it is normal to adjust with one gun, then fire one round from each gun to verify the mean point of impact (MPI) of the battery
• One battery was tasked to conduct the mission on T1 and the outcome would be circulated to the other battery
• One battery fired six rounds on target T1. The assessed actions are;
  • A Register Mission with six rounds fired with one gun adjusting (on for line / long bracket / short bracket)
  • A Register Mission  with one gun adjusting  obtaining a target round after two rounds then each gun firing one round each to verify the mean point of impact (MPI) of the battery
  • A Witness Point Mission with a single gun firing to ascertain that the rounds were falling in the desired place (two rounds) then fire one round from each gun to verify the mean point of impact (MPI) of the battery
•  As a result of the firing on the target T1 a correction is available to be used to apply the correction to  the map data for targets T2 and T3 to predict the firing data to engage those targets.
• The fire missions against targets T2 and T3 were conducted during the hours of darkness.
• Night firing unobserved would make this effectively harassing fire.