[gap]

Originally Posted by vdr1981

Randomize distance between units in groups from 800-1000m (I'm planing to increase this to 800-1100m, before was always 900)

Found some interesting info on this topic:

Distance between ships was generally about 500 feet [152.4 m] fore and aft and 1,000 feet [304.8 m] abeam, but such instructions often proved purely theoretical. [I suspect the figures above are to be intended in yards rather than in feet. If it was so, we would get some more reasonable distances of 457.2 m between ships in the same column, and 914.4 m between two adjacent columns]

source: https://books.google.it/books?id=TgF...IVQzgaCh1HiQHP

Normally a convoy was formed in a rectangular shape, with a much wider frontage than depth. Ships most commonly occupied nine to eleven parallel columns, each averaging five ships. Both weather conditions and the need to avoid collisions could affect the formation. The distance diagonally across a convoy of 45 ships might be 8000 yards [this figure is consistent with a spacing of 5 cables / 926.6 m between columns, to 4 cables / 741.28 m between ships in a column]. Escort vessels were about 3000 yards [2,743.2 m] further out in order to stand a chance of detecting U-boats before they came within torpedo range of the merchant ships. Thus the perimeter to be defended might amount in some cases to 60,000 yards [54,864 m], or 30 sea miles.

source: http://www.militaryhistoryonline.com...ic/convoy.aspx

Convoys of between 40 and 60 ships would steam in columns with 2 miles between each column [3,704 m; this figure is inconsistent with what is stated below by the same source. I suspect the 2 miles figure provided here to be a misprint for 1/2 miles, i.e. 926 m] and a third of mile [617.33 m] between each ship. A twelve-column convoy would extend 5 1/2 nautical miles in length [10,186 m; i.e. 926 m between each column] and almost two miles deep [3,704 m, if this figure was to be taken to the letter, supposing 5-ship columns (60/12 = 5) it would entail a distance of ca. 926 m between ships in the same column; again, this is inconsistent by large with the figure of 1/3 miles / 536 m stated above. Since number of ships in each column is not stated explicitly, and the figure is provided as approximate, I would rather ignore it]

[...]

In the early months of the war, naval escorts for outgoing convoys from the British Isles could go only 300 miles out to sea before having to turn back to escort incoming convoys. Escort destroyers could stay out safely for only seven days without refuelling, which meant three and half days' cover for a convoy and three and half back. This was increased to 400 miles by October 1940 and to halfway across the Atlantic by April 1941. Since air cover for shipping could also be provided from the British Isles, from Canada and Iceland, the Atlantic space left open to the U-boats was reduced by May 1941 to a width of around 300 miles. Even so, it was a yawning chasm, and in any event it did not stop German submarines exploiting the tragets wherever they might find them.
[Though not strictly relevant to the discussed topic, I decided to include this paragraph because this is an important aspect of the convoy system which unfortunately can't be fully replicated in game. A possible workaround to simulate the mid-Atlantic gap could be splitting each convoy set within GroupTypes.cfg files into 2-3 legs, one of which totally unescorted. This would entail the first leg of each convoy despawning somewhere in mid Atlantic, and a similar convoy re-spawning at the same point but without escorts. It would be a lot of work though ]

source: https://books.google.it/books?id=2yS...convoys&f=true

Experience in the First War had shown that massed shipping, on an ocean crossing, could be protected better when spread over a broad front than in two or three long lines. In the Second World war a typical convoy of, say, 44 ships would be allocated positions in columns on a grid whose corner numbering would be 11, 111, 114 and 14.

[...]

The distance to be mantained between columns was generally 1000 yards [914.4 m] and the distance between ships in a column was intended to be 800 yards [731.52 m], but this might vary according to circumstances.

[...]

Allocation of ships to specific columns would not be haphazard: groups of ships with a common destination (called 'portions') would be allocated places in the convoy appropriate to their direction of travel from the 'Split Position' on the further side of the Atlantic Ocean.

[...]

Within these portions, considerations would also be given to the value and urgency of the cargo of each ship. The materials of the highest value were, of course, oil and munitions for keeping up the fight. Food, wood and iron-ore seem not to have been given high priority, and often featured on the exposed wings.

[...]

The average speed of a fast convoy (ON westward-bound, HX eastward-bound) was nine knots an that of a slow convoy (ONS westward, SC eastward) was seven knots.
From the meeting point with the ocean escort to the dispersal point on the opposite side of the ocean might take 8 days with a fast convoy and twelve with a slow. These journey-tomes might have to be extended due to adverse weather conditions.

source: https://books.google.it/books?id=aWy...convoy&f=false

U-boat Countermeasures - October 1940
Thirty-five attacks on U-boats or supposed U-boats were made during the month. Eighteen of these by surface ships, four by submarines and thirteen by aircraft.
This month has been devoted to developing and improving measures taken to counter day and night U-boat attacks on convoys. To this end changes have been made both in the disposition of convoys and of their escorts.
Great efforts are being made to equip all convoy escorts with apparatus, which will enable them to locate a U-boat on the surface at night outside visibility distance. As more ships are fitted, dispositions will be changed so as to make the maximum use of this aid. [This appears to be the first mantion of radar being used and it is strange that a name has not been attached to the new aid.]
This new equipment has also been fitted into aircraft of Coastal Command and Fleet Air Arm. It will detect U-boats on the surface up to a range of 15 miles, and will be especially valuable for detecting U-boats on the surface at night.
It is hoped that depth charges will be carried by the aircraft, but in any event the enemy can be forced under and kept submerged until the arrival of Asdic-fitted ships, or until he is compelled to surface to re-charge batteries.
The transfer of all convoy routes to the North-Western Approaches in July led to considerable difficulty in the provision of air escorts for convoys, owing to the lack of adequate landing ground facilities in that area; these difficulties are being surmounted and it is now becoming possible to operate aircraft at night for convoy escort work. It is intended to provide the maximum air escort for three hours before darkness falls, as this is the period in which U-boats take up their positions, pareparatory to night attack.
The high percentage of hits recently obtained by U-boats in night attacks has made it necessary to increase the distance apart of convoy columns from three [555.96 m] to five cables [926.6 m]. [...] This materially reduces the theoretical chances of more than one ship being hit by a salvo.
[...] the following measures are employed by convoy escorts against the enemy tactics.

By Day
On an attack being made, escorts on each side of the convoy form on line of bearing on the mean line of advance of the convoy, ships 2,500 yards apart, then move to a position three miles in rear of the convoy, turn outwards together 90° and sweep in line abreast for one hour. All ships drop depth charges on the initial turn and further depth charges every two miles of the sweep. If by this time no contact has been obtained, escorts re-join the convoy.

By Night
Escorts are disposed in line ahead 3,000 yards apart at visibility distance from the convoy. In the event of attack, all ships on the engaged side, or on both sides if the side of attack is in doubt, turn 90° outwards together and proceed at full speeed for a distance of 10 miles from the convoy, firing star shells to illuminate the area. Destroyers are stationed in van and rear positions where possible.

source: https://books.google.it/books?id=ApB...cables&f=false

The high percentage of hits obtained by U-boats in night attacks made it necessary in November 1940 to increase the distance apart of convoy columns from about 600 yards [548.64 m] to about 1000 yards [914.4 m]. The distance between ships in the same column was about 400 yards [365.76 m]. In December 1940 the distance between columns during the daytime was reduced back to 600 yards to increase protection against air attacks.

To counter the heavy losses suffered as a result of the night attacks by surfaced U-boats in September and October 1940 the escorts were stationed in positions down each wing at a greater distance from the convoy than before. In the event of an attack, they were instructed to proceed outward from the convoy for a distance of 10 miles at full speed, firing star shell to illuminate the area where the U-boat might be, in an attempt to sight her or force her to submerge, thereby improving the chances of Asdic detection. If contact was made, two escorts were to hunt the U-boat, while the remainder were to rejoin the convoy.

Later, when radar-equipped escorts became available, they were stationed one on each beam of the convoy, about 4 miles from it in order to avoid back echoes from the convoy on the radar set. The beam escorts were to steam on the same and opposite courses as the convoy, zigzagging as requisite for self-protection. Another method of sweeping, which was under trial in order to effect an economy in fuel, was for the escort to start a slow turn of 360° when in a position abeam of the leading ships, thus sweeping outwards and astern at about 1° per second, and on completion assuming station abeam of the rear ships. The remainder of the escorts were disposed as before, but were instructed to bear in mind that, at night, the rear wing positions were the most important and that, in the event of a U-boat attack, star shell searches were to be made in the rear of the convoy also.

source: http://www.ibiblio.org/hyperwar/USN/...-51/ASW-2.html

A convoy of forty-five ships, says Roskill, would cover 5 square miles of sea [this figure is compatible with the ones provided further below by Macintyre]. A broad front was preferred; Roskill's diagram shows a 45-ship convoy disposed in nine columns each of five ships, and a 55-ship convoy in eleven columns of five. Macintyre, on the other hand, referring to the 35-ship SC 7 [October, 1940], says:

a convoy of 35 ships would be formed up in five columns of four ships and three columns of five ships each, the longer columns being in the centre. The columns would be five cables [926.6 m] (half a mile) apart, the ships in columns three cables [555.96 m] (600 yards) apart...
Such a formation was necessary for several reasons. Long columns invariably became strung out. Whereas a column of five ships would probably be one and a half miles long, 15 ships would rarely be less than six or seven. Thus the broad-fronted formation was more compact from the point of view of escorting, while signals from the Commodore, leading one of the centre columns, could be seen simultaneously from all the ships. Furthermore, it offered a smaller target to submarines which normally attacked from the flank.

source: https://books.google.it/books?id=hzP...cables&f=false

[...] convoy WS6 [April, 1941], carrying troops round the Cape of Good Hope to the war in the Middle East, was 'formed in 10 columns, three ships to a column. Speed of convoy 9 knots. Distance between columns 1000 yards [914.4 m], distance between ships in the column 600 yards [548.64 m]. Station keeping between columns fair, between ships in column, fair to poor.'

source: https://books.google.it/books?id=rra...convoy&f=false

Convoy HX-150 [September 1941], with 44 merchantmen underway from Halifax on September 16, 1941, was the first to use US escorts, which relieved Canadian escorts 350 miles east of Halifax on September 17. [At last! This is the date I was looking for, when US escorts to be added to the "FreeAmerican" roster should start spawning within Allied convoys ] The escort screen was commanded by Captain Morton L. Deyo in Ericsson. Convoy was disposed in nine columns, with distance between columns set at 600 yards [548.64 m]. Ericsson's station was 2,000 yards [1,828.8 m] directly ahead. The other four destroyers, a mix of one Benson class and three 4-pipers, patrolled 500-2000 yards [457.2-1,828.8 m] from the outer ships in the convoy. Column distances were tightened at night. On clear nights the destroyers continued their patrols, but on foggy nights they were to "keep station".

source: http://www.daileyint.com/seawar/seawar3.htm

The Countermeasures Review for August 1942 states:

[...]

Numerically our escorts are weak, but there is little doubt that technical improvements, leading to a greater reliability in such aids as H/F D/F and RDF, combined with lack of experience on the part of the U-boats, are telling, heavily in our favour.

[...]

When the ships are so fitted for reception, the first indications of the presence of U-boats have been in every case H/F D/F bearings, and these have also been of great assistance to Escort Commanders in appreciating the subsequent situations. Type 271 RDF has proved a most efficient detector by night and the quick action taken by escort vessels has, in many cases, thwarted the night attack.
This has apparently forced the U-boats to abandon their usual tactics and to attack submerged in daylight, accepting lack of mobility and a low speed against an escort weakened by memebers searching on D/F bearings. These day attacks have resulted in as many as five ships being torpedoed by one salvo; to lessen the chance of such success recurring, instructions were issued in the third week of August to open out the distance between the columns of a convoy to five cables [926.6 m] by day as well as by night.

source: https://books.google.it/books?id=ApB...onvoys&f=false

[Convoy PQ18, September, 1942]
Each column 4 cables [741.28 m] apart. Distance between ships in each column 2 cables [370.64 m] apart. Normal cruising speed 8 knots.

source: http://www.russianarcticconvoymuseum...colmEckert.pdf

The convoy [HX 212, October-November, 1942] had 45 ships, sailing in 4 rows, 12 columns, the Commodore saying that "the convoy was formed at my special request in 12 columns. I am firmly convinced that the safest formation for a convoy is when formed with a wide front, and thin in depth. Thus my arrangement for 36 ships would be in 18 columns of 2 each. No difficulty was found in manouvering the larger number of columns, and signals in practise pass outwards quicker than they do from front to rear, as the more ships in the column, the greater the distance between ships, owing to bad station keeping. Guides on the whole keep pretty good station. I would like my next convoy to try a wide front. It also has the advantage that stragglers can be seen at once from the Commodore's ship and checked.

From the escorts' point of view, the wide front should be popular, as the vessels stationed on the beam could be closer in, and have less area of water to cover. The front of the convoy, if ships were say 3 to 4 cables apart, could be covered by 3 vessels, and one would be required at least astern. An extra destroyer would be useful for reconnaissance at a distance in advance and astern of convoy. In my opinion, this last is one of the most important of the escorts' duties, when air reconnaissance is not available".

Distance between columns: 700 yards [640.08 m].
Distance between ships in columns: 500 yards [457.2 m].
Average speed: 8.69 knots.

source: http://www.warsailors.com/convoys/hx212.html

The convoy [HX 221, December, 1942-January, 1943] had 38 ships.
Average speed: 8.6 knots.
Distance between columns: 1000 yards [914.4 m].
Distance between ships in columns: 500 yards [457.2 m].

source: http://www.warsailors.com/convoys/hx221.html

Enroute Casablanca, French Morocco to United States. Convoy G.U.F.-3 [December, 1942 - January, 1943] consisting of 9 ships escorted by the U.S.S. NEW YORK abd U.S.S. PHILADELPHIA and screened by DESTROYER SQUADRON 15. C.T.F. 35 [...]. Convoy is formed in 4 columns, distance between columns 700 yards [640.08 m], distance between ships in columns 500 yards [457.2 m]. The U.S.S. NEW YORK, convoy guide, is leader of second column. U.S.S. PHILADELPHIA is stationed 1,000 yards, twenty five degrees on starboard bow of the U.S.S. NEW YORK. DESTROYER SQUADRON 15 is stationed on a semi-circle from approximately 8,000 yards on each beam of the guide to 6,000 yards ahead of the guide.

source: http://www.brigs.us/phila/Gardner/19..._Diary0224.pdf

[...] the convoy [ONS 5, April-May, 1943] was arranged in 12 columns with 4 ships in each, except for column 6 which had 2 and column 12 which had 3 ships. The distance between each ship in the column was set at 800 yards [731.52 m], with 1000 yards [914.4 m] between each column, making the front of the convoy 5,5 miles wide and a little over 1 n. mile deep.

source: http://www.warsailors.com/singleships/bonde.html

The orders for the convoy [KMS 18B, June-July 1943] specify a cruising speed of 9 knots and a distance of three cables – 600 yards [548.64 m] - between ships in columns

source: http://davidgibbins.com/journal/2014...y-10-july-1943

[Convoy HX 246, June-July, 1943]
Distance between colums - 1000 yards [914.4 m]
Distance between ships in columns - 500 yards[457.2 m]

source: http://www.warsailors.com/convoys/hx246.html

Next convoy to be subject to fancy Luftwaffe tactics was UGS-37 [March-April 1944] consisting of 60 merchant ships and six LSTs escorted by five United States destroyers, eight DEs, two British PCs, two tugs and aintiaircraft cruiser H.M.S. Delhi. [...] The convoy was in close formation, 400 yards' [365.76 m] distance between ships, 600-yard [548.64 m] intervals between columns, making 7 1/2 knots over a calm sea [...]

source: https://books.google.it/books?id=1BL...onvoys&f=false

The largest mercantile convoy to sail in this or any other war was HXS 300, which sailed New York on 17 July 1944 and arrived U.K. on 3 August with 167 ships and only 7 mid-ocean escorts. With 19 columns, this convoy had a front of some 9 miles [16,668 m; i.e. a distance of ca. 926 m between columns].

source: http://www.ibiblio.org/hyperwar/USN/...-Convoy-3.html

CONVOY No. O.N.270 [December, 1944]

Distance between columns:
East of 7 deg West 3 cables [555.96 m]
West of 7 deg West 5 cables [926.6 m]

Distance between ships in column:

To R/V 2 cables [370.64 m]
Than 3 cables [555.96 m]

Normal cruising speed: 9.5 knots

source: http://jproc.ca/rrp/apend_b2.html

summing up, these are the possible combinations (distances rounded to the nearest multiple of half a cable, i.e. 1/20 nmi, ca. 93 m):

Column		Row
spacing (m)	spacing (m)	Convoy(s)
===============================================================================
556		371		UGS-37 (Mar-44), ON-270 (Dec-44)
556		?		HX-150 (Sep-41)
649		371		none
649		463		HX-212 (Oct-42), GUF-3 (Dec-42)
741		371		PQ-18 (Sep-42)
741		463		none
741		556		none
927		463		HX-221 (Dec-42), HX-246 (Jun-43)
927		556		SC-7 (Oct-40), WS-6 (Apr-41), ON-270 (Dec-44)
927		741		ONS-5 (Apr-43)
927		?		HXS-300 (Jul-44)
?		556		KMS-18B	(Jun-1943)

Personally I would use a random mix of the combinations above, with a preference for larger spacings (especially the 927 by 556 m formation) from October '40, and even more from August 1942. Also note that the shortest spacings need to be tested, as ships in game might tend to scatter in panic of collision, when steaming in close-formation convoys.