Dive time in v1.3 screwy?

[Tobus]

I noticed yesterday that my Tambor dives better when lying still (0 kts) then when under speed. This seems strange to me, since when diving under speed uses the diveplanes, where as diving at zero speed only utilizes the ballasttanks.

Did a timing with my tambor (calm weather) from surface to PD:
speed = 0kts: 25 seconds
speed = 9 kts: 38 seconds

Shouldn't this at least be the other way around?

[chopped50ford]

IT should be...the momentum of the ship and diving planes forcing the ship downward should make it dive quicker.

I would be curious to see what the response is on this one. :up:

[Suicide Charlie]

When I first started looking at the SHIV side of these forums I was browsing through the screenshot thread a lot. I noticed a large amount of shots showing the downing of aircraft. In SHIII, I almost ALWAYS dove as soon as I spotted aircraft as did most other players that shared their stories.

When I finally got SHIV I figured out why. The subs dive terribly slow, even when crash diving. It like your only choice is to stay up and fight instead of rendering yourself so defenseless as you wait for your boat to sink below the waves.

[hyperion2206]

When I first started looking at the SHIV side of these forums I was browsing through the screenshot thread a lot. I noticed a large amount of shots showing the downing of aircraft. In SHIII, I almost ALWAYS dove as soon as I spotted aircraft as did most other players that shared their stories.

When I finally got SHIV I figured out why. The subs dive terribly slow, even when crash diving. It like your only choice is to stay up and fight instead of rendering yourself so defenseless as you wait for your boat to sink below the waves.

This was true until patch 1.3 was released. Now you always get a warning when aircrafts are near you giving you plenty of time to crash dive.

[Suicide Charlie]

Radar warning, yes. You're usually able to dive before they get on scene and in a bombing run.

[joea]

US subs did take longer to dive but how long? Compared to U-boats?

[Rockin Robbins]

In wartime, there was no casual "take her to periscope depth." EVERY dive was a crash dive and if desired they caught the sub at periscope depth as best they could. If you're just pushing "p" to dive, the real sub jockeys would laugh at you. There's only two kinds of subs: those who crash dive and those that die.

In practice, the gold standard was 60 seconds to periscope depth. Dives were practiced several time per day and if a dive was performed in over 60 seconds, woe unto the crew, for they would drill mercilessly until they could meet that standard. Good skippers tolerated grumbling crews who thought they were overdrilled because grumbling crews are still alive. A little action and a freighter or two in Davie Jones' locker will fix that just fine.

[JSF]

I find the dive time in acceptable limits now with 1.3. Big improvement over earlier model.

[Tobus]

Guys, focus.

I'm not making assumptions about which sub should dive faster than the other, how that was done in a purely technical sense, or if divingspeed is altogether correct.

I am merely stating that in the current state in v1.3 (haven't tested this in earlier versions), divetime is much shorter when speed is 0 compared to when the sub is at speed. In my opinion, shouldn't this be the other way around?

Since a moving sub has added advantage of water flowing over the diveplanes as the sub is moving in the horizontal plane. A non-moving sub should therefor change depth much slower than a moving sub.

In SH3 (+GWX ofcourse), this is the case. In SH4, it seems reversed.

[tater]

Moving should absolutely be faster. That's why a crash dive rings up flank. Subs in general don't seem to use the dive planes at all. All dives seem level instead of being steep.

tater

[SteamWake]

Funny, I thought the S boats dived ridicously fast and the others 'felt' right.

But its just a 'feeling' I have no real life comparisons.

[nomad_delta]

Agreed, all subs should dive much faster at flank than when stationary. That's what the dive planes are for -- translate speed into downward momentum.

If they're not behaving accordingly, that's a serious problem. I think I'll add this to my investigation "to-do" list... ;)

nomad_delta

[Steeltrap]

I've posted a number of times on this subject of dive behaviour. I even went so far as to test time taken to dive/rise in 50' brackets from surface to 250' and back at various speeds, from ahead 1/3 to flank.

My conclusion was this:

The dive times are hardcoded. You take x seconds to go from 50 to 100, y from 100 to 150 etc. Those times remain unaffected by your forward speed, which is totally unrealistic. The dive angle is also purely 'visual' in effect. At slow speed this is apparent as you travel more quickly vertically than horizontally, while the reverse should always be true unless you are using excessive blowing/pumping. Look at your stern planes and then look at the pitch of the sub - the two are unrelated. Changing depth once already submerged involves your stern remaining near horizontal with the stern planes on full dive/rise.

Just another 'realism' thing that's out the window.....

[Tobus]

Agreed, all subs should dive much faster at flank than when stationary. That's what the dive planes are for -- translate speed into downward momentum.

If they're not behaving accordingly, that's a serious problem. I think I'll add this to my investigation "to-do" list... ;)

nomad_delta
Thanx Nomad. I'll see if I have the time to test more subs and post my findings here. I'm a mod-noob, so I don't think i'll look into any of the files.

[nomad_delta]

Ok, I've run some tests, and here's the data I've come up with so far. All tests were performed on a clean install of SH4 1.3 with no mods applied. All timing was done w/ a real stopwatch (not the one in the game) and no time compression. All commands were issued via the keyboard shortcuts, not by clicking on the depth meter. So here's the data:

http://img53.imageshack.us/img53/8191/sh413divetimeskk0.png

Here's where it gets really interesting: you'll notice in the chart there I have some of the dive times marked as "DRY" and some marked as "WET". There may be a proper term for this that I'm unaware of, but in my chart "DRY" specifies that the submarine has been surfaced for at least 20 minutes, while "WET" specifies that the submarine has been underwater very recently -- within the last couple of minutes.

You can see in the chart there that DRY vs WET makes a huge difference in overall dive times.

It took me quite a while to figure out what was happening, but it appears that it takes approximately 20 minutes for the water to "drain" out of the limber holes in your sub after you surface. You can actually watch the water draining out of the sides of the sub -- it will stop completely after about 20 minutes. If you dive again before 20 minutes is up, you'll be able to dive much faster than you would otherwise. It's not an on/off thing, either -- there's a gradual progression between the WET/DRY dive times I show in the chart during that 20 minutes.

---

I don't have enough time to write up a detailed analysis of the dive times and speeds right now, but at first glance it appears that Tobus is right -- something is screwy, even once we have the 'wet' vs. 'dry' thing accounted for.

What's odd is that it's not consistent: the 'dry' normal dive times look right with Flank being faster than Stop, while the 'dry' crash dive times are about the same for Flank and Stop.

The 'wet' dive times are what really don't make any sense. Here we're seeing exactly what Tobus was talking about: both subs dive faster at "All Stop" than they do at "Flank" speed.

Another interesting thing I've learned from this is that there really is a difference between "Crash Dive" and normal "Dive" that goes beyond the fact that "Crash Dive" rings up Flank for you. Same goes for "Emergency Surface" as opposed to normal "Surface" -- definitely faster.

----

I plan on collecting the same data for the remaining subs and I'll update the chart once I'm done; should be pretty interesting, I think.

nomad_delta

(edit - almost forgot something else I found that's probably important: I tried all of the above tests both a Battle Stations and not, and saw no significant difference between dive or surface times at all!)

[Tobus]

WOW, that's one great test!:o

Screwy kind of physics the dev's have used on this one.

[FooFighters]

Wow.. this is really getting interesting..
Strange diving times devs :-?

[kv29]

people here never cease to amaze me, flaws can´t hide anywhere :up:
are the involved numbers tweakable?

[SteamWake]

It has been known for some time that if there is a short period of time between dives you will dive faster.

The assumptioin is that there is still 'water' in the ballast tanks so they dont have to 'fill' all the way.

The dive planes evidently are just eye candy. In the previous versions they dident even come down half the time and had no effect on dive times. Unfortunate but true.

The 'data' chart above is about what I expected to see.

You want to see some really diverse results throw a sugar boat into that data.

[simonb1612]

thanks for running the tests. It seems strange that the devs would consider such things as amount of residual water in/on boat when calculating dive times but then ignore the fact that the earth is spherical when doing travel distances (see the thread about fuel efficient routes)
http://www.subsim.com/radioroom/showthread.php?t=118877

[nomad_delta]

It has been known for some time that if there is a short period of time between dives you will dive faster.

The assumptioin is that there is still 'water' in the ballast tanks so they dont have to 'fill' all the way..

I didn't, but I sure wish I'd known that before I started testing. It would have saved me a whole lot of time & banging my head against the wall trying to figure out what I was doing wrong when my numbers weren't making any sense at all. :D

It does make total sense, now, though. The devs even bothered to include the "water draining out of the sub" animation which lasts exactly as long as the effect on dive times dos, and the dive times even change gradually as the water drains out. I think that's pretty awesome.

Does anyone know if there is a 'more correct' term to describe the ballast tanks being full of water or not than what I've used above? I didn't know what to call it.

The dive planes evidently are just eye candy. In the previous versions they dident even come down half the time and had no effect on dive times. Unfortunate but true.

They're not quite just eye-candy, as they do seem to have an effect in some of the tests, just always the correct effect (as Tambor said, depending on the circumstances it can even be backwards) and not as much of an effect as I would think they should.

The 'data' chart above is about what I expected to see.
You want to see some really diverse results throw a sugar boat into that data

I plan to, actually; Now that I've got my methodology down it shouldn't take quite so long to run the tests, and I'm hoping we'll have a clearer picture (for better or worse) once I've added the data for the other subs.

As soon as I'm done I'll put a a new chart with all the info.

nomad_delta

[nomad_delta]

people here never cease to amaze me, flaws can´t hide anywhere :up:
are the involved numbers tweakable?

As soon as I'm done testing, I'm going to find out. :p

nomad_delta

[JREX53]

Not all of the water is blown out of the ballast tanks when the sub first surfaces in RL. The subs I was on had a High Pressure blower to blow out the residual water in the ballast tanks. It does take about 20 minutes in RL to get all of the water out of the tanks. So this might have been factored into the game, but I don't know for sure.


edit: That should have said a "Low" Pressure blower, not "High" Pressure Blower.

[SteamWake]

Not all of the water is blown out of the ballast tanks when the sub first surfaces in RL. The subs I was on had a High Pressure blower to blow out the residual water in the ballast tanks. It does take about 20 minutes in RL to get all of the water out of the tanks. So this might have been factored into the game, but I don't know for sure.

Well keep in mind were talking about real life physics vs a computer 'model'.

My guess is that somewhere there is a set of values for the ballast tanks for the sake of example lets say 0 = empty 100 = full.

When you broach the surface they are probably somewhere around 60 and decreases over time to emulate 'draining'.

Just a guess.

[JREX53]

Not all of the water is blown out of the ballast tanks when the sub first surfaces in RL. The subs I was on had a High Pressure blower to blow out the residual water in the ballast tanks. It does take about 20 minutes in RL to get all of the water out of the tanks. So this might have been factored into the game, but I don't know for sure.
Well keep in mind were talking about real life physics vs a computer 'model'.

My guess is that somewhere there is a set of values for the ballast tanks for the sake of example lets say 0 = empty 100 = full.

When you broach the surface they are probably somewhere around 60 and decreases over time to emulate 'draining'.

Just a guess.

Isn't that what I said. I know we are talking about a Computer 'model' not real life.

[-Pv-]

"...every time hit the dive key blow crush depth it'll rise to crush depth..."

Try diving from deck awash (22ft at the keel and diesels still running) and you'll have the supports under water in about 15 seconds. This is the profile I use for all my surface attacks as well as defending against aircraft.
-Pv-

[Canonicus]

I've posted a number of times on this subject of dive behaviour. I even went so far as to test time taken to dive/rise in 50' brackets from surface to 250' and back at various speeds, from ahead 1/3 to flank.

My conclusion was this:

The dive times are hardcoded. You take x seconds to go from 50 to 100, y from 100 to 150 etc. Those times remain unaffected by your forward speed, which is totally unrealistic. The dive angle is also purely 'visual' in effect. At slow speed this is apparent as you travel more quickly vertically than horizontally, while the reverse should always be true unless you are using excessive blowing/pumping. Look at your stern planes and then look at the pitch of the sub - the two are unrelated. Changing depth once already submerged involves your stern remaining near horizontal with the stern planes on full dive/rise.

Just another 'realism' thing that's out the window.....


Well....They aren't....You can change the rate at which the ballast tanks fill-up by changing a value in the DATA/SUBMARINE/NSS_*/NSS_*.sim file using the Minitweaker app. Category 4 is called flooding.

You'll see this...

ManBT_flood_speed = 20000 ( or a similar number)

Changing the number to read, lets say, 35000, will cause the tanks to flood faster and consequently give you faster dive times.

For example...

- ManBT_flood_speed on Tambor/gar/gato from 20,000 to 35,000
(crash dive speed increases from 1min 10 seconds to 46 seconds)

- ManBT_flood_speed on Salmon/Sargo from 20,000 to 29,000
(crash dive speed now in the 45-48 second range)

Another file to look at is called...

DATA/SUBMARINE/NSS_*/NSS_*.zon
------------------------------
Adjust the crash speed value from 2 to 1.


If you don't know how to use the MiniTweaker app...then use Ducimus's Trigger Maru 1.3 mod, which has these adjustments already made for you.

Cheers

[Rockin Robbins]

How are you measuring dive times? In RL dive times were from the order to dive until periscope depth was reached. Less that 60 seconds was the goal.

40 to 48 seconds would be unrealistically fast.

[Steeltrap]

Hi Canonicus.

I'm not trying to make the boat dive more quickly. I was - and perhaps didn't explain it well - making the point that your dive times from one depth to another (be it surfaced to 50', or 50-100, 150-200 etc...) seem to be fixed, irrespective of forward motion. This is most definitely an 'error' as subs of those times were particularly reliant on forward motion for depth changes, in combination to minor pumping/blowing of trim tanks.

I tested this at 1/3, 2/3, standard, full and flank speeds. What I found was that the times taken to change depths were almost identical at ANY speeds (although not the same for different brackets).

In other words, if it took 20 seconds to go from 100' to 150' at 1/3 speed, it took almost exactly the same amount of time to do this at flank. When I said 'hard-coded' I meant that these times for changes in depths seem to be fixed. I should have chosen another term. Interestingly, it took longer to change depth by 50' when you were deeper than at shallower depths, even if you weren't coming from the surface. It also tended to take longer coming UP compared with going DOWN.

One question: does changing those flood rates have any effect on depth change rates once you are already dived? i.e. does it only affect rate of diving from surfaced state? What about from 150'-200' if you've been at 150' fro 30 minutes?

Either way, it seems to me that the current situation of being able to change depths at the same rate at any speed is clearly unrealistic/flawed.


Cheers

[tater]

How are you measuring dive times? In RL dive times were from the order to dive until periscope depth was reached. Less that 60 seconds was the goal.

40 to 48 seconds would be unrealistically fast.

From what I've read 30-35 seconds was the goal. 60 was a prewar standard time. Silent Running talks about drilling the crew for 30-35 second crash dives and specifically states that slower than that is dangerous. Similar in other books I've read.

tater

[Canonicus]

Hi Canonicus.

One question: does changing those flood rates have any effect on depth change rates once you are already dived? i.e. does it only affect rate of diving from surfaced state? What about from 150'-200' if you've been at 150' fro 30 minutes?

Either way, it seems to me that the current situation of being able to change depths at the same rate at any speed is clearly unrealistic/flawed.


Cheers

Well... I guess the only way to know would be to input the higher flooding rates and then go do a little testing.

Is it unreasonable to think that if you were to make the ballast tanks flood faster that it also might have the effect of making them blow faster, as well?

I do know that with the higher flood rates in place,you go down much faster from the surface. As far as the effect on sub-surface depth changes and the time it would take..I'm not sure.

Again...it would need to be tested.

Would be interesting to see if it does.

[nomad_delta]

Just realized I didn't post this along with the chart:

All "dive" times are time from surface to 120 feet.
All "surface" times are time from 120 feet to 22 feet.

For the dives I didn't have the sub stop at 120 but rather continue past it as fast as possible to avoid having it slow down to level off at that depth so I would get consistent measurements.

Same for the surface times -- I stop the clock when the sub reached 22 feet, as each sub takes a while to level off once it reaches the top and I didn't want that affecting the measurements.

Mostly what I'm after are the dive times relative to eachother, to make sure they make sense in context -- which they don't.

nomad_delta

[chopped50ford]

When your diving, you have to remember the air has to escape the ballast tanks through small holes and the positive bouyancy has to reverse to the negative to get any momentum downward. It takes time. In game, I normally crash dive then come back up. It seems to be quicker in time...but I havent put a clock on it either.

Im my experience as a real Helmsman/Planesman :rock: ...on both early and late 688 Subs, there was a huge "dive time" difference between "Sail Plane" boats to the dive planes on the bow.

If my memory serves me right, the bow plane boats submerged much faster because of the help of both sets of planes (stern planes/bow planes) working together. In the "Sail Plane" boats, the sail planes did not come into full effect until the decks were completely awash with 10' of the sail under water.

I know that there is a huge difference here (especially in boat types), but i do recall everytime we dove, it was down first then back up to PD...never from the surface directly to PD. It was hard enough to keep it 5-10 feet below the surface w/out broaching.

We should be glad that PD in this game is pretty consistant. Our Helmsman/Planesman are awesome since (unless your in a sea state 3-5 storm) the top of the boats never broach the surface of the water.

Imagine if they put in a "chief of the watch " feature and you had to pump or remove ballast from front, rear or center to keep the boat level. In a sense, it wouldnt be a bad idea. Then you can pump out ballast especially if your taking in water from damage.

These boats have normally (1) pump...the newer boats had several...each controlled by the C.O.W.

[Tobus]

Wow, my guess really stirred up a hornet's nest.:up:

Great reading and testing guys, maybe something can be modded of this by you smart guys.

[hyperion2206]

Here's my impression: I started in '43 with a Gar class and green crew and diving seemed to take at least 60 seconds. Now, at the end of '43 I've got a Balao class sub and the lowest rank on board is Petty Officer 3rd class and now the sub dives in about 45 seconds.
So I assume that the rank and experience of the crew influences the dive times. Can anybody confirm?

[BH]

Diving times in TM 1.3a are very quick.

[-Pv-]

I can't confiirm with hard testing, but my impression on watching my porpoise dive times improve over the course of two years of war patrol, is the efficiency of the crew is a factor.
-Pv-

[supposedtobeworking]

So how can we tweak diving times if we want to but don't use TM 1.3a?

[PepsiCan]

Hi

I think we should slightly focus the question. It is not as to how fast a sub should dive and it is also not about whether a crew's efficiency leads to better dive times.

The primary point uncovered through Nomad_delta's excellent method of test-and-prove, is that the relationship between speed and dive times is inverse!

And it is clearly wrong, given logic and given reality, that boats dive slower when they have more speed.

So, it seems that there should be a variable in the SIM files somewhere that got reversed or maybe the code in the executable reverses the interpretation of the variable in the SIM file. Either way, it's a bug and it is the second one Nomad_delta has uncovered since 1.3 came out (the Contact/Influence switch not working being the first).

I think we need to ask for a patch 1.4.