Ballast tanks:

Sorry for the delay. I meant to post this stuff last week, but had second thoughts about the results. That's happened more than a few times; what looks good at 3 AM when you are tired and want to finish, doesn't always look ok in the harsh light of day. :-?



I did supplemental tests with the S, Tambor, and Narwhal classes, changing only the flood rates of the ManBT and DiveBT tanks.

In each case, I ordered the crew to GQ, and got the boat up to 10 kn. Ordered a dive to 165 ft. and recorded the time at 35 and 40 ft. Then I surfaced the boat and TC'd to get the boat back to a normal equilibrium, and ordered a 'crash' dive and recorded the times again. The reason I used 35 and 40 ft. was that this would be just enough to allow the full submergence of the vessel, without the dive planes or other factors making a large difference. The flood times I chose were based on provisional values I had decided on, and reasonable spreads above/below those. (14 sec. for Tambor, 42 sec. for S-class, 56 sec. for Narwhal.)

Flood rates were calculated as follows:

(submerged disp. - surface disp.) / (flood time * .00114) = ManBT

.00114 converts from liters/sec. to tons/sec.


S-18 class,

ManBT 4795, DiveBT 48
[30 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:32, 0:42 ________Crash ________0:22, 0:26

ManBT 3425, DiveBT 35
[42 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:34, 0:48 ________Crash ________0:23, 0:27

ManBT 2664, DiveBT 27
[54 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:37, 0:54 ________Crash ________0:23, 0:27



Tambor class,

ManBT 43860, DiveBT 439
[10 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:27, 0:30 ________Crash ________0:22, 0:25

ManBT 31328, DiveBT 314
[14 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:32, 0:37 ________Crash ________0:26, 0:30

ManBT 24366, DiveBT 244
[18 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:36, 0:42 ________Crash ________0:29, 0:33



Narwhal,
ManBT 18170, DiveBT 182
[42 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 1:00, 1:15 ________Crash ________0:40, 0:49

ManBT 13628, DiveBT 136
[56 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 1:18, 1:35 ________Crash ________0:44, 0:55

ManBT 10902, DiveBT 109
[70 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 1:30, 1:55 ________Crash ________0:47, 0:59



With the S class, starting with a 30 sec. flood time we get to 35 ft. in about 32 sec., as we might expect, but adding 12 sec. to the FT adds only 2 sec. to the result, and adding 24 sec. to FT adds only 5 seconds. Even stranger is the fact that the crash dive times hardly change at all!

The Tambor seemed to behave more rationally. Starting with a 10 sec. FT gives a time of 27 sec. Adding 4 sec. to the FT adds 5 sec. to the result, and adding 8 sec. to the FT adds 9 sec. to the dive. Crash dive times were only slightly faster starting at 22 sec. and going up 4 and 7 seconds, respectively.

With the Narwhal, a FT of 42 sec. produces a dive time of 60 sec. and adding 14 sec. to the FT adds 18 sec. to the result. Adding 28 sec. to FT adds 30 sec. to the result. Crash dive times start at 40 sec. and go up by 4 sec. and 7 sec.



To try and clarify the results I did a few supplementary tests:

I repeated the Tambor tests (18 sec. FT) but lowered the DiveBT from 244 to 122, and then to 12. The results were the same. The DiveBT value seems very unimportant in all this. I will come back to this in a little bit.

Here is a retest of both the S, and Narwhal classes using values to give them 14 sec. flood times like with the Tambor:



S-18 class,

ManBT 10275, DiveBT 103
[14 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:26, 0:31 ________Crash ________0:19, 0:22


Narwhal,
ManBT 54511, DiveBT 545
[14 sec. flood time]

Dive_______time at 35, 40 ft. _____ Dive _____time at 35, 40 ft.
Normal_______ 0:31, 0:36 ________Crash ________0:27, 0:31

We see that the Narwhals times are very close to the Tambor, but the S-class is considerably faster, perhaps because it is smaller.

So, it is apparent, that crash dive times are much faster than normal dive times, and this is especially so where normal dive times are slow. ManBT values affect normal dive times more than crash times and DiveBT values are of little import, so long as they are above zero. The only time I noticed any change from the DiveBT is when I used zero. The boat seemed to dive normally to a deep depth, but when I ordered p/s depth, the boat went above and could not get back down to p/s depth.

I just decided to use 1% of the ManBT values. Whatever the mass/buoyancy factor of the 'dive' tanks, they cannot be changed and we are stuck with them. It isn't clear to me if this is intended, or is another part that was miscoded.

Note that there is a benefit obtained with crash dives, even if there is no propulsion. I wasn't sure about this until I tested it. So one can order crash dive, and then reduce speed immediately to stay quiet.

I would've liked to use real/historical values of flood rates/ times, but I could not find anything on this. In my mind, historical ratings for dive times are somewhat murky, due to the fact that we have very little detail in terms of the procedures used. For instance, in the game MBT's start flooding seemingly at once, but I'm not sure that this should be so. We can mod the flood rates to delay complete filling, but cannot delay the start of the process. It is hard to believe that the S, and Narwhal classes had tank arrangements that flooded that slowly, yet we know they were considered to be very slow divers. For these reasons, diving performance is likely to be a thorny issue.

Here is a preview of USN sub performance.

TAMBOR CLASS

Acceleration from stop:

RFB
speed, kn. ___5_______10______15______20

time, sec. ___0:05____0:11_____0:25____1:07


ISP 2.0
speed, kn. ___5_______10______15______20
time, sec. ___0:21____0:43_____1:15____2:25



Turing, 180 degree turn:

RFB
surface at 15 kn.
slows to ________ time to complete_________dia. of circle
6.0 kn._______________4:16________________520 yd.

submerged at 6 kn.
2.3 kn.______________11:53 ________________ 550 yd.

ISP 2.0
surface at 15 kn.
9.8 kn. ______________2:20 _________________500 yd.

submerged at 6 kn.
4.8 kn. ______________4:38 _________________450 yd.



Diving from 10 kn. to periscope depth:

RFB
depth, ft. ____20 ________40 ________60
time _______0:07 ______0:32 ______ 2:15
speed, kn.___9.0 ________6.6 _______ 5.2
angle, deg. _ - __________ -1 _______ 0

ISP 2.0
depth, ft. ____20 ________40 ________60
time _______0:08 ______0:38 ______ 1:36
speed, kn.___9.0 ________4.8 _______ 4.5
angle, deg. _ - __________ -6 _______ -0.5



Diving from p/s depth to 200 ft. @ 6 kn.:

RFB
depth, ft. ____80 _____120 _____160 ____ 199
time _______0:27 ____1:25 ____2:58 ____ 8:05
speed, kn. ___5.7 _____5.5 _____5.3 _____5.9
angle, deg. __-1 ______-2.5 _____-3 ______0

ISP 2.0
depth, ft. ____80 _____120 _____160 ____ 199
time _______ 0:31 ____1:16 ____2:11 ____5:24
speed, kn. ___5.3 _____5.2 ______5.1 _____6.0
angle, deg. ___-6.5 ____-10 _____-9.5 _____-0.5


[Climbing from 200 ft. to p/s depth is comparable, with ISP angle reaching about +7 deg. and times being faster for both.]



Crash dives from surface starting at 10 kn.:

RFB
depth, ft. ____40 ______80 ______120 ______149
time _______0:26 ____0:51 _____1:31 ______3:55
speed, kn. ___10.8 ____10.2 _____9.2 _______9.0
angle, deg. __-6 _______-5.5 _____-5 ________0

ISP 2.0
depth, ft. ____40 ______80 ______120 _____160 ____199
time _______0:29 ____0:54 _____1:15 ____1:39 ____3:45
speed, kn. ___7.5 _____7.9 ______8.0 _____7.9 _____9.0
angle, deg. ___-7 ______-13.5 ____>-15 ____>-15 ___-0.5


Battery performance is unchanged from v1.1.


In both RFB and ISP, the subs have a much easier time going from deep submergence to p/s depth than vice-versa. By this I mean that the last 10 feet or so of the depth change is much faster going up than down. It is not clear to me why this should be so, but it is. I worked a great deal trying to minimize this slow creeping tendency, but was only partly successful.



Other fleetboats have similar performance. As I said before the Gatos seemed to behave differently for no good reason; the dive plane drag coefficients had about double the effect, so I halved the numbers and adjusted so they would come out about the same.

The S-class have slower flood times for the MBT's and have more difficulty getting under, but are calibrated to take on a 10 degree dive angle as the others.

The Narwhal are very sluggish and unmaneurverable (by design), and have very slow flood times. They are also more 'loggy' once under the surface.



The S-class still have turning circles of ~800 yds., but nevertheless benefit from making faster turns. I gave the Narwhals turning circles of about 650 yds. (no data for these).

Note that I used propulsion numbers from after their '42 refits, so they are not so underpowered as they would be before. Still, they are not the same as the fleetboats!



Next post:

Surface Ships

Surface Ships:

'max_force' and acceleration

The key to reducing the acceleration of ships was getting the 'max_force' values right. This was accomplished by carefully tabulating enough results, and changing one input at a time to permit a good analysis of the results. There are too many ships to try to do individual trial and error work here.

In S3D, the 'max_force' values are said to be in tons, but I knew this wasn't correct, early on. If it was the case, large ships would require larger values of 'max_force', than I used. In every case, I sought to use values at, or very near, the minimum required to achieve the stated maximum speed of the vessel in question.



I skip some of the mind-numbing detail here, but I first came up with a formula based on purely empirical observations. However, I wasn't fully confident of this, so I did a series of more systematic experiments, and analysis. This lead to the same equation, except that the results regarding what I called the potential maximum speed were not clear-cut, and possibly should be to a power greater than 1, like S^1.2 or S^1.3. I tried this, but the simpler formula I had before seem to produce the better results, so I used that.

F = K * P / (M * S)

where

F is max_force (??)
P is eng_power (effective horse power)
M is mass (tons)
S is maximum speed (knots)
K is a constant

I used this formula to calculate the values of 'max_force' for both subs and ships alike, and tests showed that results were very acceptable. By this I mean craft were able to reach maximum speed, but with much slower acceleration than before.

For the ships, I took the displacement, at either 'full load', or 'normal load', and used half the rated engine hp. I can't remember where I read it, but the source considered the effective horse power to be about half the rated engine horse power. This seemed right to me. Higher eng_power makes for faster acceleration, and faster deceleration. Lower values reduce both. I didn't have any specific figures to go by, but this seemed to work well. [For instance, if you are submerged, you should be able to slow from 6 kn. to 3 kn. reasonably fast without backing engines.] Then I used the above with a value of 0.15 for K to obtain the proper value of 'max_force'.

The same procedure was used for the subs, then the calculation was repeated using the submerged propulsion numbers; displacement, eng_power, and max_speed, with a value of 0.20 for K. The higher result of the two was retained. Usually, the subs had more difficulty getting up to their max speed when submerged.

For example, the Clemson DD has a powerplant rated at 27,600 hp and a displacement of 1,308 t, and a maximum speed of 35.5 kn.

Half of the value of 27,600 hp is used, in the calculation, to obtain a 'max_force' of 0.0446. Note that half of 27,600 hp is also used in the *.sim file.

Ok, now you're probably wondering what does this 0.0446 number really mean. What does this come to in real money?

Well, it turns out that the maximum acceleration is approx. equal to 1000 times the 'max_force', in knots per minute. So, for the Clemson, we can expect that it would be able to reach about 11 knots in 15 seconds. Of course, drag reduces the rate of acceleration as the speed increases; this is the maximum.



Turning Performance

I experimented quite a bit with different number schemes as far as rudder drag and prop factors are concerned. There aren't necessarily clear-cut advantages of one over another.

RFB and stock use higher rudder drag and low prop factors. This slows down subs turning quite a lot. I decided in the end to use a higher prop factor and lower drag. This made for a big improvement in the subs, but about the same results for surface ships. This was somewhat disappointing, but perhaps is a realistic result. (I couldn't find much of the way in data for this speed drop.) In any case, I found a rudder drag of 0.01 and prop factor of 0.5 to give good results for most ships. For very tight turning ships, like DE's, I used 0.015 and 1.0, respectively. This was necessary, as the lower numbers, even when combined with lower LR drag numbers, didn't cut it. IMO, the stock prop factors of 0.05 are just too low, and most of the turning effect should come from the prop factor, with only enough of rudder drag to balance things out. If no rudder drag is used, ships tend to overturn and it gets very sloppy.

For LR drag, I used 0.2 for most ships, but 0.1 for some tight turning ships, and 0.3 for destroyers that turned poorly.



Stability

I didn't make many changes in this area. It didn't seem to be needed, for the most part. Someone recently mentioned problems here in RFB, and I did happen to see instability in the Fubuki DD while testing. In ISP 2.0, it seemed ok, though the GC (gravity center) was the same. I suspect this is due to small differences in the draft. If anyone finds they need to correct instability, I would try adjusting draft first before messing with the GC. Small increases in draft seem to help quite a bit.

I did adjust the draft/GC of the small minesweepers and kaibokans in the mod, as I remembered seeing them rolling back and forth.

Weather and Sea-keeping

I tried mightily to get realistic sea-keeping for ships, like with the subs, but no go. Most of the speed loss with the subs in heavy seas comes from the game computing increased drag when the hull goes down into the water. The game has to do this to have sensible diving behavior, of course. Only the smaller part of the speed loss is due to the screws coming out of the water. With the ships, there is a small speed loss with the screws coming out of the water, but this doesn't last long enough to have a large effect. There is no defined submerged propulsion for the ships, so there is no increased drag effect either.

I mentioned some time ago, I was able to produce wind speeds above 15 m/s in missions. Unfortunately, in a campaign, the weather would stick, and would not change. So I reluctantly set this idea aside, and left the weather as it is in v1.1. I may try to develop some better weather mod down the road.



dom111 mentioned the severe speed drop in merchants that do the constant helming thing. I considered reworking the turning in such a way as to specifically counteract this tendency, but decided against it. From a physics standpoint, I think the turning is pretty good. It is probably possible to reduce this by reducing the rate of turning, but it is not clear to me that it would be a worthwhile tradeoff. In any case, I thought it better to get feedback on things as they are. Perhaps, I might revisit this issue in the future.





Below are a short sample of ISP 2.0 ship tests:

Hog Island freighter [max_force .003780]

RFB acceleration from stop
speed _____0 ______2 ______4 ______6 ______8 ______10
time _____0:00 ___ 0:04 ___ 0:08 ___0:15 ___0:27 ____0:49

ISP 2.0
speed _____0 ______2 ______4 ______6 ______8 ______10
time _____0:00 ___0:29 ___0:59 ____1:33 ___2:17 ____3:29



New Mexico BB [max_force .004080]

RFB
speed ______ 0 ______5_______10 ________15 _______20
time ______ 0:00 ___0:07 ____0:19 ______0:41 ______ 1:33

ISP 2.0
speed ______ 0 ______5_______10 _______15 ________20
time ______ 0:00 ___1:05 ____2:17 _____3:52 _______6:58


Brooklyn CL [max_force .018900]

RFB
speed ____ 0 _____5_____10 _____15 _____20 ____25 ___30
time ____0:00 ___0:04 ___0:05 ___0:08 ___0:13 __0:20 __0:36

ISP 2.0
speed ____ 0 ______5_____10 _____15 _____20 ____25 ___30
time ____0:00 ___0:15 ___0:29 ___0:45 ___1:04 ___1:31 __2:16



Clemson DD [max_force .044600]

RFB
speed __0 ____5 ____10 _____15 _____20 ____25 _____35
time __0:00 __0:03 __0:04 ___0:05 ___0:07 __0:09 ____0:29

ISP 2.0
speed __0 ____5 ____10 _____15 _____20 ____25 _____35
time __0:00 __0:08 __0:14 ___0:20 ___0:27 ___0:38 ___1:23



In the mod, I'm putting in the options folder, a test mission where you can launch torpedoes under some sitting ships, to see how they accelerate. The mission should be installed manually, so you can run it with, and without the mod, to see the difference here.

Acceleration and Deceleration


One of our members mentioned that his boat didn't respond well when stopping or backing. (This in RFB.) This didn't surprise me as I had done some tests early on with ISP.

Of course, I wanted to post some specific numbers. I figured it was easier to run a new test, than to try to find the relevant notes from months ago (if I even made any).



submerged, at 6 kn., All Stop

RFB
speed, kn. ____6 ______5 ______4 ______3 ______2
time _______0:00 ___0:48 ____1:45 ____3:21 ___6:18

ISP 2.0
speed, kn. ____6 ______5 ______4 ______3 ______2
time _______0:00 ____0:15 ___0:33 ___0:57 ____1:46


submerged, at 6 kn., Back Slow

RFB
speed, kn. ____6 ______5 ______4 ______3 ______2 ______0
time _______0:00 ____0:40 ___1:34 ____2:34 ___3:59 ___5:43
[can reach 2.6 kn.]

ISP 2.0
speed, kn. ____6 ______5 ______4 ______3 ______2 ______0
time _______0:00 ____0:11 ___0:25 ____0:40 ___0:59 ___1:21
[can reach 3.0 kn.]

The striking difference here is due to the low value of E propulsion eng_power used in RFB and stock. This illustrates why it is undesirable to rely on manipulation of eng_power alone, to dampen acceleration.

Note, that acceleration characteristics are comparable.




Ships

I decided to use the RSRDC selection of ships. I believe most RFB users usually use RSRDC, and RSRDC adds a number of ships to the game. I didn't want to have to come up with two separate mods and I don't think ISP 2.0 will work properly without RSRDC. The extra ships won't do any harm, but RSRDC added aircraft and changed air groups of carriers (and maybe some cruisers?), so I had to work from RSRDC files.

There were a few ships in RSRDC which did not have *.sim files. This bothered me, as I felt that they should have distinctive propulsion characteristics. These included the Yubari light cruiser, some destroyers, and maybe a few others. Others, like the Clemson APD, and Momi Patrol Boat, received major changes, since their capabilities seemed to harken back to their destroyer days.

Ok, I'm not sure what the circumstances are.

Did this happen in a mission, a career, in the open sea, in a harbor?

If you are talking about ships in port, know that docked ships cannot move - period. That has nothing to do with the mod. Ships that are simply standing still, but not docked, should be able to accelerate according to their engine/propulsion details. Freighters are mostly underpowered, and will be slow to get up to speed.

If you are asking about torpedoed ships, damage will likely slow them down. I haven't changed the damage model.

Uninstalling ISP 1.1 and installing ISP 2.0 is ok, but I wouldn't do it in the middle of a combat situation. I have no idea what might happen.

I am really not clear what happened here.


***

Thanks, aanker.

:subsim:

Ok, I get the picture.

The only thing I changed in the damage values is lowering the crash speed in the *.zon file. I did this so that the ships would not be sinking in heavy seas. A fatally damaged ship should still sink; just not as fast. Weapons damage and ship hit points were not changed.

If I get more reports of this nature, I'll consider changing the crash speed back. I'm reluctant to do this, however, as this could result in most ships sinking in stormy weather. The game options in this area seem to be limited.

Thank-you for taking the time to report back.

I'm starting work on v2.1 now.

To get a better idea of the severity of the problem, I set up a mission and shot torpedoes into different merchant ships.

Medium Old Composite Fr.

no ISP
1333 hit mid., 1344 hit bow, 1355 going down, 1357 gone

ISP 2.0
1318 hit mid., 1319 hit stern, 1429 hit bow, 1557 hit mid., 1800 2 hits bow stern, 1838 2 hits mid, bow, 1838 going down, 1840 gone



Medium Modern Split Fr.

no ISP
1336 hit mid., 1422 hit stern, 1446 going down, 1447 gone

ISP 2.0
1322 hit mid., 1432 hit stern, 1617 hit bow, 1617 going down, 1622 gone


Medium Modern Composite Fr.

no ISP
1339 hit mid., 1407 going down, 1409 gone

ISP 2.0
1326 hit mid., 1511 hit stern, 1625 hit stern, 1640 hit bow, 1711 hit mid. - did not sink



Small Old Split Fr.

no ISP
1454 hit mid., 1557 hit stern, 1602 going down, 1604 gone

ISP 2.0
1844 hit mid., 1942 hit stern, 2015 hit bow, 2055 hit mid., used 50 + shells - rammed twice - won't sink.

without ISP, I was able to 7 other merchants with 10 other torpedoes (Mk. 14). This was all with auto targeting and no torp malfunctions. With ISP 2.0, I ran out of torps after the above four ships.


I also used the deck gun to destroy a fishing boat with ISP 2.0. It only took about 10 rds. and registered sunk quickly, though it seemed to take longer to go down completely.

It would seem that the "normal" flooding is interrupted by the changed CrashSpeed, and sinking targets required either some sort of critical damage, or blasting away all the target's hit points ??

Odd that some targets will seem to sink easily, but often not.

Yes, that is basically true.

However, many ships were left in their stock configuration. A mod, Warship Damage Model was released to be used with RFB, with more ships being altered, but I don't think they ever got to all of them. So some are in one kind of configuration, and some are in another.

Here are some samples:

Stock
survivor ............................1 hit point
CMD_small_boat ..................15
JPFish01 ............................50
most merchants ....................300 to 400
most BB's .............................800 to 1000
Yamato BB ...........................1500

Naka CL ................................2160 [somebody has a high opinion of this one! :)]


RFB 2.0
CMD_small_boat ..................50
JPFish01 ............................4000

most merchants ....................4000
most DD's ............................4000
US subs ...............................15000

RFB doesn't change most cruisers, battleships, or allied DD's. So one is left with the oddity of merchants having more hit points than a BB, and a sub having much more again.

My understanding is ships can be destroyed in 3 ways: 1. all hit points are exhausted, 2. destruction of 'critical' area (like keel), 3. be pulled down by flooding. I guess the last is what usually happens with RFB.



If anybody has this mod, Warship Damage Model, could you upload it to subsim?

Last time I tried, I had the same problem.

When I have time I'll try to find it in some of my external HD's. I thought I saved everything from my old drives, but I haven't found it yet.

***

Since the modified *.zon files in ISP 2.0 are causing problems, I did a test with some ships with those files. I was pleased to find a group of 9 different merchants was able to continue at 9 kn. through 12 m/s seas without sinking for 24 hours. This is with my bigger waves, of course. I will have to test some of the other ships, but this is encouraging.

Since I will have to make a new version, I think I will try to fix (or reduce) the 'constant helming' problem.