WesWorld

To avoid turning Wes's Atlantean ship thread into a huge mess..

Back in 2013 I looked into some of the flaws of simming submarines. One of the main things being the "flat" submarines being simmed which to me is caused by two contradictory things that the springstyle notes tell us to do. The first being "You will sim your sub in awash condition, just about to go under" and the second being: "Specify depth as about 2/3 of beam; this seems typical for subs of pre-nuke era." The first bit tells us that we need to sim it in such a way that the draught is the height of the submarine from keel to main deck, but when I look at cross sectional images of submarines, I do not get the impression at all that the depth is about 2/3 of beam as the notes suggests as being "typical". Far from it. They look the be more like 1:1 in most cases. Looking at line drawings of a number of subs, I get the next:

Type VII depth is 97% of beam.
Type IX depth is 88% of beam.
Type XXI depth is 113% of beam.
I-201 depth is 121% of beam.
Gato class depth is 86% of beam.
Balao class depth is 100% of beam.
D-class depth is 74% of beam.
E-class depth is 77% of beam.
V-class depth is 100% of beam.
N-class depth is 101% of beam.
G-class depth is 79% of beam.
K-class depth is 84% of beam.
L-class depth is 89% of beam.
X1 depth is 98% of beam.
USS Holland depth is 115% of beam.

The closest to the depth = 2/3 beam is the British D-class submarine but even then its depth is 74% of the submarine's beam which still isn't anywhere near the 66.667% that the springstyle notes tells us is "typical for subs of pre-nuke era". To me 80-100% seems much more typical for submarines than 66.667%.

Some may remember these two pictures I posted some time back, one comparing the I-305 beam and depth with Australia's K-2 class and the other comparing a ("more proper" since the first one was considered to look too modern) I-305 with the Type VII and the I-201 cross sections.





Looking at the sims of existing and former submarines in the encyclopedias regarding the beams and depths (note: most of this list was made back in 2013 and I quickly tried to add the subs that have appeared after I made that list so I might have missed a few)...

Atlantis
D-1 class 58%
D-9 class 58%
D-22 class 62%
R-26 class 80%
R-38 class 79%
R-62 class 55%
R-74 class 55%
R-98 class 77% (subsim)
MSS-1 class 66%
Type 42 class 60% (subsim)
Type 43 class 61% (subsim)

RSAN
C class 55%
D class 69%
E class 60%
F class 62%
G class 63%
H class 70%
I class 65%
J class 57%
K class 67%
L class 66%
M class 64%
N class 66%
O class 65% (not enough misc for ballast)
P class 67%
Q class 60%
R class 67%
T class 66%

Argentina
Salta class (1931) 60%
Capitan Valles class 60%
Salta class (1947) 111% (subsim)

Brazil
Tubarao class 60%
A class 67%
B class 63% (subsim)
C class 70%

Canada
H class unknown (no sim)
L class unknown (no sim)
Shearwater class 67%
Orca class 64%

Chile
Capitan O'Brien class 62%
Union class 79% (subsim)
Gualcolda class 63% (subsim)
Lautaro class 56% (subsim)
Heroe class 89% (subsim)

Mexico
R class 67%
D-22 62%
Delfin class 62%
XSS-1 class 66%
R-26 class 80%
Remora 61%

Peru
Type P class 60%
SC Class 60%
P-6 class 55%

USN
R Class 67%
S Class 68%
Porpoise Class 77%
Gato Class 67%

Australia
H class 90% (data=?)
A class 66%
B class 62%
J class 60% (data=?)
K Class 52%
SSH Class 62%

China
Typ I 58%
Type IIc 80%
Jin Class 62%
S Class 50%
E Class 60%
Guiyu Class 67%
Shayu Class 79%

Chosen
Incheon-Class 62%
IO-V class 63%
IO-VII class 59%

India
IX-1 class 65%
I-2 class 67%
I-6 class 68%
I-9 class 67%
I-11 class 71%
I-15 class 65%
I-21 class 65%
I-22 class 65%
I-24 class 71%
I-28 class 72%
I-31 class 80%
I-34 class 57%
I-36 class 69% (subsim)
Durjeya class 74% (subsim)

Japan
Shinobi I class 67%
Shinobi II class 67%
IC-I class 59%
IM-I class 67%
IO-I class 67%
IO-II class 60%
IO-III class 63%
IO-IV class 67%
IC-II class 55%
IC-III class 61%
IC-IV class 65%
IO-VI class 59%
IO-VIII class 71%
I-300 class 121% (used upper belt for ballast instead of miscellaneous weights to get this percentage)
I-305 class 81% (though I would like to sim it with a higher percentage)
Sangadila 66%
I-310 class 107% (subsim)
Ha-100 class 107% (wikistats)
I-402 class 78% (subsim)
I-314 110% (subsim)
I-315 class 107% (subsim)

Persia
I-2 class 67%
I-5 class 60%

Philippines
Remora class 60%
Delfin class 66%
Trucha class 66%
Agulon class 65%
Aluhaman class 79% (subsim)
Mako class 64% (subsim)
Audaz class 79% (subsim)
Tirador class 77% (subsim)

Thailand
Perla Class 87%
Dolphin 64%
J class 60% (data=?)
Dolphin II 76%

Bulgaria
Delfi class 55%
Leshtanka class 63%
Uzhilvane class 73%
Akula class 79% (subsim)

Denmark
D Class 63%
Gemini 42%
F Class 58%
Aegir Class 62%
Bellona Class 65%
Ue Class 55%

France
Eagle class 66%
Protee class 65% (subsim)
Roland Morillot class 60% (subsim)
CS-27 class 50%
CS-30 class 63%
Daphne class 79% (subsim)
Thetis class 79% (subsim)
Roland Morillot-class 60% (subsim)
Gymnote-class 90% (subsim)
Emeraude-class 67% (subsim)

Germany
Type I 60%
Type IV 64% (subsim)
Type VII 63% (subsim)
Type IX 77% (subsim)
Type XXI class 67% (subsim)

Greece
Nemesis Class 70%
Delphinos Class 61%
Delphin Class not enough data
Hermes Class 72%

Iberia
Sa Class 66%
Sb Class 63%
Sc Class 66%
Sd Class 65%
Se Class 67%

Ireland
Bradan Feasa class 87% (subsim)
Pooka class 84% (subsim)

Italy
S Class 50%
B Class not enough data
SX.01 66%
SA Class 65%
SB Class 71%
Brin Class 71%
Da Vinci Class 86% (not enough misc for ballast)
Antonio Sciesca Class 84%
Argonauta class 80%
Adua Class 89%
Perla Class 87%
T Class 83% (subsim)
U Class 64% (subsim)

Latvia
Krovona Class 75% (SS3)
Baltica Class 73%
Afrika Class 65%
Continental Class 83% (subsim)

Nordmark
S Class 70%
H Class 71%
G Class 57%
J Class 70%

Netherlands
K.1 Class 56%
O.1 Class 72%
K17 Class 67%
Zwaardvisch Class 60%
K-35 Class 66%
O-8 Class 66%
O-17 Class 51% (subsim)
K 37 Class 51% (subsim)

Poland
P class 59%
G class 57%
A class 50%

Romania
Perla Class 87%

Russia
Kovarniy Class 67%
Bars class 63%
Delfin class 58%
Bezmolvniy Class 70%
Khitriy class 72%
Molchkom class 63%
Bezmolvniy_II Class 62%
Besshumniy class 57%
Modified Besshumniy class 58%
Delfin class 58%
M401-class 94% (subsim)

Turkey
H class 78%
H-boat (Group Three) 67%
Type I class 60%

RN
Shark Class 54%
Thames Class 54%
T Class 75% (subsim)
O Class 61%
P Class 61%
U Class 63%
S Class 69%
X Class 100% (data=?)
V Class 61%
S Class (Batch II) 69%
W Class 81% (subsim)

Yugoslavia
Type I 60%


To me the problem with these "flat" submarines is not just that what the springstyle notes tell us. Another big issue is that if springstyle/springsharp is being used and going for the more realistic depth = 80-100% beam while using the miscellaneous weights for the ballast tank. Doing this will result in stability dropping below 1 a lot quicker (because miscellaneous weights in SS2 and older would be floating somewhere above the water with the freeboard at 0 instead of being inside the submarine) meaning that you would be pretty much at the ballast tank being 1/6 of the normal displacement and barely any miscellaneous weights for other things such as schnorkel, radar, sonar, torpedoes, etc.

The stability issue is the main reason why I use other things in the sim for certain weights of those extra objects (deck armor and belt armor) to slow down getting to that stability limit. It is also the main reason why I used the upper belt weight for ballast instead of miscellaneous weights when I tried to sim the I-201-like I-300. The only way to get close to the I-201's depth = 121% beam with springsharp while properly using miscellaneous weights would be to ignore all the stability issues since springsharp dumps the miscellaneous weights in a totally unrealistic position for a submarine anyway. But then since we are pretty much brainwashed by the idea that stability needs to be above 1 with submarines which means that if we are to add more miscellaneous weights to a design, we will need more beam to solve the stability issue.



Now with that behind me there are a few other things (non-dimensional relater) that I have noticed:

- Submarine speed and range
With SS2 sims based on historic subs, I had this habit to use the speeds and ranges of those historical submarines and work from there. However, as mentioned above, the springstyle notes tell us that the sub is simmed in "awash condition, just about to go under". It came to me one day that that would mean that the speed and range I enter are actually the speed and range of the submarine in "awash condition, just about to go under" and not the same as the surfaced values that are given in stats for those historical submarines... meaning that I always have wasted a lot of time trying to match sim speeds and ranges with those of historical submarines.

- Construction
Another thing that I more recently thought of (and not back in 2013) is this: How much of a submarine is actually finished when it is at 40% of construction? Looking at some launch pictures, the submarine needs to be fairly complete in order to be launched so my answer would be that at 40% a submarine will not be complete enough to be launched, not even close. I will use Japan's new I-315 as example. Now considering that with launch the submarine is pretty much closed off, it needs to have all the machinery (diesel and electric) inside as well as all the batteries. Also probably needed are the torpedo tubes, the air-powered piston ejection pumps, Ninjatousaya System, sonar some of the 'camouflage', and the improved control surfaces. All other things I think can be added later. So removing all the unneeded stuff would give me a light displacement (according to subsim) of about 6100 tons that is needed to be in the submarine before it can actually be launched... and that is 92.6% of the submarine's light displacement and not 40%. Sure KISS is the easy way and we are pretty far underway with the sim to change things, but this completely unrealistic aspect of submarine construction is kinda annoying to me.


Edit:
Another comparison picture of below water cross sections of the Japanese submarines. The top and middle line of simple images give an idea how the older Japanese simmed subs look like with the black circle being the pressure hull based on the depth of the design and the brown box around it gives the extremes of the beam compared to the pressure hull. The more detailed bottom row are the 1940s submarines where I ignore the incorrectness of the "depth = 2/3 beam" given in the notes. I would prefer the I-305 to be more like the I-300 and I-310, but since I simmed the I-305 using the legal way to sim a submarine in SS2 (with miscellaneous weights) I ended up with more beam than I wanted. With subsim I would have done it differently.

Of the top and middle line, only the I-400 would be okay since that one would internally be similar to the cross section of the I-402.

Quoted from "Rooijen10"

Back in 2013 I looked into some of the flaws of simming submarines. One of the main things being the "flat" submarines being simmed which to me is caused by two contradictory things that the springstyle notes tell us to do. The first being "You will sim your sub in awash condition, just about to go under" and the second being: "Specify depth as about 2/3 of beam; this seems typical for subs of pre-nuke era." The first bit tells us that we need to sim it in such a way that the draught is the height of the submarine from keel to main deck, but when I look at cross sectional images of submarines, I do not get the impression at all that the depth is about 2/3 of beam as the notes suggests as being "typical". Far from it. They look the be more like 1:1 in most cases. Looking at line drawings of a number of subs, I get the next:

I suspect most people look at the L/B/D figures of historical submarines and then use that to crib in their own dimensions for subs; but then the "sim in awash conditions" is forgotten, and people neglect to add a few feet or meters that would normally be above the surface for the L/B/D conditions.

For most of the subs I've designed, I generally use the historical L/B/D numbers for a historical boat, and only diverge in order to satisfy the "max 11:1 length-to-beam ratio" rule we have for ship designs.

Quoted from "Rooijen10"

Another thing that I more recently thought of (and not back in 2013) is this: How much of a submarine is actually finished when it is at 40% of construction? Looking at some launch pictures, the submarine needs to be fairly complete in order to be launched so my answer would be that at 40% a submarine will not be complete enough to be launched, not even close. I will use Japan's new I-315 as example. Now considering that with launch the submarine is pretty much closed off, it needs to have all the machinery (diesel and electric) inside as well as all the batteries. Also probably needed are the torpedo tubes, the air-powered piston ejection pumps, Ninjatousaya System, sonar some of the 'camouflage', and the improved control surfaces. All other things I think can be added later. So removing all the unneeded stuff would give me a light displacement (according to subsim) of about 6100 tons that is needed to be in the submarine before it can actually be launched... and that is 92.6% of the submarine's light displacement and not 40%. Sure KISS is the easy way and we are pretty far underway with the sim to change things, but this completely unrealistic aspect of submarine construction is kinda annoying to me. :)

I considered this myself when I started playing. My response is generally to put all of the tonnage down in the first quarter (or two, if necessary), launch the boat once it's mostly applied, and then just keep them under construction at dockside for the remainder of their build time, as the shipfitters do the rest of their tinkering.

Quoted

I suspect most people look at the L/B/D figures of historical submarines and then use that to crib in their own dimensions for subs; but then the "sim in awash conditions" is forgotten, and people neglect to add a few feet or meters that would normally be above the surface for the L/B/D conditions.

I think that when you look at Wikipedia for example, those figures are actually that of a surfaced vessel.

Using the Type VII as example, it is given on wiki as having a beam of 6.20 m and draft of 4.74 m. It also states a height, but that most likely includes the conning tower. Now based on those wiki stats, that would mean that the depth is 76% of the beam, but as indicated above, the line drawing actually indicates that the depth is 97% of beam for a Type VII in awash conditions and not 76%.

That means that when you use the stats from a historical submarine, you already start wrong, which becomes even more wrong when you start playing with the numbers, especially when you are forced to add beam when ending up with stability less than 1. But you do not think about that because "Hey, don't the springstyle notes say that depth = 2/3 of beam which is typical for pre-nuke era submarines?"

The I-201 stats are a bit more reliable to use for sims as it actually gives "Height: 7 m (23 ft) (keel to main deck)" instead of a surface draft figure (which, looking at other language pages, appears to be 5.46 m). it is that height which you need to know for the "sub in awash condition, just about to go under" bit. It is also that bit that got me thinking back then whether the depth is indeed 2/3 of the beam.

Quoted

For most of the subs I've designed, I generally use the historical L/B/D numbers for a historical boat, and only diverge in order to satisfy the "max 11:1 length-to-beam ratio" rule we have for ship designs.

As far as I'm concerned, that applies to waterline length (and I believe I have mentioned before that is used to be 12:1 before it was changed for some unknown reason and without any discussions to 11:1). With submarines, since you "sim in awash conditions", the length of the submarine is the length (oa) and not length (wl). So in a sense that rule tells us that it is okay to have a ship with a length (oa) : beam ratio that is greater than 11:1 as long as the length (wl) : beam ratio (which SS2 gives us) is 11:1 or less but it is not okay for a submarine to have length (oa) : beam ratio that is greater than 11:1.

The I-201 stats suggest a l:b ratio of 13.62 and since the I-300 is based on the I-201 I used that ratio because it's a historical design, even though the gentleman rules say I am not allowed to do so.

Quoted

I considered this myself when I started playing. My response is generally to put all of the tonnage down in the first quarter (or two, if necessary), launch the boat once it's mostly applied, and then just keep them under construction at dockside for the remainder of their build time, as the shipfitters do the rest of their tinkering.

Good to know that I am not the only one thinking that way.

Your way of doing it is not a bad idea actually for submarines. If the decision is made to stop Wesworld once we get to 1950 and start Wesworld 2, I think that that is probably something that should be looked at when it comes to submarines.

... maybe since we are in a fairly early stage with the current version of Navalism, it is probably something for me to bring it up there as well...

I generally look at real world dimensions when designing subs but I've never really taken conscious notice of the awash condition in the springstyle notes. I agree though that stats can be suspect and measuring draught on a submarine when surfaced must miss a few metres above water. Like Walter I recently had that moment of realisation and my latest SubSim designs have a couple of metres added when based from real world dimensions.

I don't ever 'launch' a submarine in my reports, if you notice I omit that completely and just give lay down and completion dates. I assume all construction is in the dock (I never use slips) for the entire building time and 'launching' is when the completed sub is floated out for her first builder's trials.

Good point on the construction aspect, my latest subs building at Palinerus basically are launched when they are 100% complete, but that's mainly because of the back story on the Yards at Palinerus being drydocks linked together in order to facilitate fast construction assembly line style. It does make sense though to launch subs at the point where they are almost complete.

Quoted from "thesmilingassassin"

Good point on the construction aspect, my latest subs building at Palinerus basically are launched when they are 100% complete, but that's mainly because of the back story on the Yards at Palinerus being drydocks linked together in order to facilitate fast construction assembly line style. It does make sense though to launch subs at the point where they are almost complete.

I am uncertain that is a justifiable assumption.

On a number of occasions I have made reference to the website US Shipbuilding History. It has a wealth of data on construction times, among many other things. I’ve looked at the tabular data for US submarines constructed during the Second World War for that data is broken down to the slip/dock time, time in water fitting out, and overall time, with the dates specified.

Assume if you will that the total time from keel laying to completion represents a linear progression of construction – which is the situation implicit in our present rules – then on average a US submarine was launched when 64% complete. The median value was 68%. In a number of cases it was less than 50%.

My point is, on average, 64% of the total time of construction was expended on doing those things to get a submarine to the condition where it could be launched, leaving 36% to the fitting out and completion.

For the sake of simplicity in our game, we have adopted the iron law of shipbuilding (9 months plus tonnage/1,000 months for ships over 499 tons) and allow a ship to be launched at 40% of completion. I think that rule has little or no basis in fact – and that can be demonstrated – but it is what we have accepted in the past and I do not argue that we change it at the present juncture. By the same token, while it might be shown that submarines were substantially more complete at their time of launch than 40%, this phase of the game is not the time to attempt to change it. If Wesworld2 ever comes to pass, it would be played using rules agreed upon by the players, whomever they might be.

Well it wasn't really an assumption on my part, I was just basically showing how I have been launching my subs as of late. Its more for simplicities sake than story line and its just coincidental that they finish closer to Walters idea of what percentage subs really should finish at. As we've seen on the discussion the general consensus is that subs launch at a higher completion percentage than what our rules allow but as Bruce says we are really too far along in the sim to make any changes and I doubt at this point its really worth while anyway. It seems to me that the percentage could vary from nation to nation but at the very least seems to be at least 20% higher than what our rules allow and perhaps even higher.

One other thing about submarines with subsim: Where does 'coastal submarine' end and 'oceanic submarine' begin?

To me, if one player sims a submarine with a 7,019 nm @ 10 knots as 'oceanic' and another player sims a submarine with a 9,818 nm @ 10 knots as 'coastal', then one of them has to be wrong.

The thing is that while the coastal submarine can't go as deep as the oceanic submarine, it can be much better in all other aspects compared to a similar sized oceanic submarine.


Throwing out a quick test with a 70x6x6 submarine with 3000 ElecHP, 2400 DieselHP, 6TT, 24 mines/torpedoes, 180 tons oil and 180 tons batteries, I get these other stats:

Oceanic:
Displacement: 1260 tons
Kerb wt: 1127 tons
light weight: 923 tons
res buoyancy: 11%

Coastal:
Displacement: 1260 tons
Kerb wt: 856 tons
light weight: 652 tons
res buoyancy: 32%


Speeds and ranges are in both cases exactly the same so I won't bother posting them.


Increasing the coastal submarine to the same light displacement of the oceanic submarine will lower the speed and range a bit, but more than enough reserve buoyancy is gained to bring those values back up (after getting the speed and range back to the original values, the test sim still ended up with a reserve buoyancy that was 14% more than what I started out with).

Getting to 915 tons light displacement (so only a few tons off the light displacement of the oceanic submarine) and bringing the speed and range back up to the above values, I now have a submarine with a reserve buoyancy of 46% which 35% more than the oceanic submarine with almost the same light displacement, same speeds and same ranges and same armament.

By accepting the sacrifice in diving depth, you can have a coastal submarine that is much cheaper (30% cheaper in my example) than an oceanic submarine with pretty much the same stats or you can have a coastal submarine that is much better than an oceanic submarine with the same light displacement. So looking at that, I really feel that it is appropriate to have a maximum range for coastal submarines and a minimum range of the oceanic submarines with a buffer between those two values.

Quoted from "Rooijen10"

So looking at that, I really feel that it is appropriate to have a maximum range for coastal submarines and a minimum range of the oceanic submarines with a buffer between those two values.

Personally, I tend to keep my 'coastal' submarines below about 700t light, and switch to oceanic if I'm going much above that limit.

I'm not opposed to establishing a guideline, but over the last century I'd say the dividing line has changed pretty substantially. A fleet submarine in 1920 might be 800 tons, whereas a coastal submarine in 1970 might reach 2,500 tons.

Quoted

Personally, I tend to keep my 'coastal' submarines below about 700t light, and switch to oceanic if I'm going much above that limit.

Yes, but as indicated with the test example values I gave above, just by changing it from an oceanic submarine to a coastal submarine the light displacement drops below your 700 ton line while at the same time you gain a lot of reserve buoyancy you can use on other stuff and, with the exception of the diving depth, all the stats are the same as when it was a oceanic submarine. Just by changing one value, you are going from 'unacceptable as a coastal submarine' to 'acceptable as a coastal submarine'. It might even be possible that, after dumping what you want on the submarine, as an oceanic submarine the reserve buoyancy drops below 0, while as a coastal submarine it is still well above the target of 10% in which case, just by changing one value, you are going from 'unacceptable as a submarine' to 'acceptable as a submarine'.

Quoted

I'm not opposed to establishing a guideline, but over the last century I'd say the dividing line has changed pretty substantially. A fleet submarine in 1920 might be 800 tons, whereas a coastal submarine in 1970 might reach 2,500 tons.

I was actually looking more at range as dividing line than displacement, although that will no doubt have the same time period issue as with the displacement you mentioned, where the average 1970 coastal submarine will have a greater range than the 1920 ocean-going submarine. Still, to me a coastal submarine does not need to have much range whereas a ocean-going submarine does need it and it also seems proper (to me at least) that the range of a submarine simmed as a coastal one should not be greater than that of a submarine simmed as a oceanic one when they are pretty much from the same time frame.