Realism- and gameplay-related hardcode fixes for SH3.EXE

[Wolfstriked]

Originally Posted by h.sie

@wolfstriked: your wish is something for an animation or graphics modder, but not for me. I wouldn't know where to start. Sorry.

Copy that.

[LGN1]

Hi Hitman,

from my experience with trying to model buoyancy I have concluded that it's a quite complicated subject. Before starting to code anything, one should have a thorough discussion in a dedicated thread about how to model it. Without any good and convincing model in mind, it would be useless to start.

Cheers, LGN1

Originally Posted by Hitman

Thanks Sharkbit It's all a matter of practicing enough

@H.Sie, did you finally shelve the idea of implementing real buiyancy, i.e. the UBoat changes depth when moving too slow for the dive planes to produce an effect?

Until now it was a compromise, either positive (GWX) or negative (NYGM) and never the two things depending on the situation as the real submarines do. But with your ability to add code maybe we could do one of two things:

Option A:

Have positive when <14 metres depth (Positive at periscope depth) and
Have negative >14 metres depth (I.e. when submerging deeper for evading).

Option B:

Have positive when normal running and
Have negative when silent running and below 2 knots, as that is a good way to reflect bilge pumps being shut off

[h.sie]

Regarding bouyancy 1: Isn't it the job of the LI to level the boat? I ask, because if we model bouyancy, then the player (the kaleun) would have to take care for the correct depth beneath all the other tasks. Wouldn't it be more realistic to rely on the LI who is more or less able to keep depth? In sh3 the kaleun often has to calculate the firing solution, send the crew to bed, and now he gets a new job: control depth. So the rest of the crew can go home.

Regarding bouyancy 2: Instead of negative or positive buoyancy: What about a continuously unsteady depth, sinosoidally deviating around the player-set-depth. The maximum difference from the player-set-depth could be dependent on 1) the LI's experience 2) the speed of the boat and maybe 3) Silent-Running. The intention behind this idea: To model the hard-working LI having more or less problems to level the boat.

Regarding fuel consumption: Would it be more realistic to enlarge fuel consumption for flank speed?

Gähn!
h.sie

Originally Posted by h.sie

Regarding bouyancy 1: Isn't it the job of the LI to level the boat? I ask, because if we model bouyancy, then the player (the kaleun) would have to take care for the correct depth beneath all the other tasks. Wouldn't it be more realistic to rely on the LI who is more or less able to keep depth? In sh3 the kaleun often has to calculate the firing solution, send the crew to bed, and now he gets a new job: control depth. So the rest of the crew can go home.

Regarding bouyancy 2: Instead of negative or positive buoyancy: What about a continuously unsteady depth, sinosoidally deviating around the player-set-depth. The maximum difference from the player-set-depth could be dependent on 1) the LI's experience 2) the speed of the boat and maybe 3) Silent-Running. The intention behind this idea: To model the hard-working LI having more or less problems to level the boat.

Regarding fuel consumption: Would it be more realistic to enlarge fuel consumption for flank speed?

Gähn!
h.sie

I like idea number 2 but my question is if your at a dead stop, will your boat bob up and down say 5 meters every so minutes, or would it be random each time you set a depth and then it would stay at that depth? (So say you dive to 20 m and since your LI's experience is low you dive to 24m instead and stay there)

and I'm not sure about fuel consumption but isn't there a mod already that gives you the right fuel economy?

[UltimaGecko]

Originally Posted by h.sie

@UltimaGecko:

Thanks for the detailed data! Your post contains so much information, that I still couldn't understand all you wrote. I'll have to read again. Your conclusion seems to be "between-the-lines" - hard to understand for me as a non-native english speaker. What is your conclusion? Do you agree with our approach or don't you?

In my opinion it is much more important to have a good game balance between the UBoat and the opponent destroyers, than to exactly fit historical data. (I remember an old thread of Stiebler regarding this topic).

What is the benefit of modelling historically accurate diving times, if in-game the destroyers endurance when hunting an Uboat is much shorter than in reality? So I agree with LGN1, that we have to fit the Uboats diving-times to the time-scale of sh3 which seems to be shorter. So we can only estimate the diving times. But this simple and qualitative approach we used (a renewable and a non-renewable oxygen component and their dependency on silent-running, torpedo reload, repairs) is a big step forward compared to stock sh3 (only renewable "oxygen" part, no dependency of diving-times at all).

Since our O2 model is only an coarse estimation and a qualitative approach instead of an exact model, it would not make any sense in my opinion to add more detailed dependencies like the crews experience on O2 consumption.

But we can surely discuss about fine-adjusting the diving-times and the balance between renewable and non-renewable part of O2.

h.sie

German is the only other language I could come close to explaining things in, but if that would be helpful I can try.

As for whether CO2 or O2 measuring is more suitable, it seems to me like CO2 is a bit more of a problem. But there are convincing arguments for both (and I probably don't have all of the available information for either side).


Affecting Both:

One thing to keep in mind is that - unlike the default game - CO2 scrubbing, oxygen renewal and ship ventilation are not instantaneous like in-game. VIIs and IXs had a pair of blowers that could efficiently move air around the ship, but it still took a few minutes to recirculate the air (so the most unrealistic aspect of surfacing, opening the hatch and having "Oxygen reserves at 100%, Captain!" is that it would take a few minutes to actually get air circulated to every part of the boat).

Second, the batteries generated a significant amount of hydrogen gas. H2 becomes inflammable around 4-5% concentration, and there's at least one uboot that suffered a hydrogen fire. So the air concentration is - unfortunately - not as simple as Oxygen, Nitrogen and Carbon Dioxide.


Carbon Dioxide:

Early in the war it seems like CO2 would be the bigger limiting factor.

The system for VIIs and IXs seems to be designed to provide about 5 days of submerged CO2 scrubbing time; this seems higher than the provided O2, but there's very little renewable CO2 time available (e.g. surfacing and circulating the air and diving really only provides you an hour or two of air before you want to start cleaning CO2 out).


There's only a short window after you dive before CO2 scrubbing needs to start to maintain a low concentration of CO2. During silent running a smaller recirculating blower (Umwälzlüfter) would be used, which had trouble providing ventilation to the entire boat (this is one of the reasons CO2 concentration can increase over 2-3% even with scrubing available).

There seem to be indications that captains would also use the smaller blower in the interests of fuel economy (as the larger, normal usage blowers ran on diesel fuel, possibly even through the Diesel engines). Enclosed spaces also tend to have higher CO2 levels. So it's possible that upon diving a crew may already have higher than atmospheric CO2 levels (of 0.4%).

In those cases - even if scrubbers are available - there may not actually be enough air moving through them to extract enough CO2 to bring the level down. Apparently the packages could be open and the NaOH spread over an area to help lower local CO2 levels (and probably increase the chance of chemical burns for whoever's doing it).


High levels of carbon dioxide can also result in acidosis. Carbon dioxide's toxicitiy is related to its partial pressure and not its concentration. This would be a compelling reason not to use compressed air to augment oxygen capacity.

For example: say you have a 5% CO2 concentration at 1 atm (sea level air pressure). This is effectively a 0.05 atm partial pressure of CO2. Say you decide to use compressed air to bring in oxygen and now there's 2atm of pressure inside the sub (the compressed air has a CO2 partial pressure of 0.04 atm). CO2 concentration is now down to 2.7%, but your partial pressure of CO2 is now at 0.054 atm (or the equivalent of 5.4% concentration at 1 atm).

(I know the in-game dial says it's measuring CO2 concentration, but it's assuming 1 atmosphere (101.325 kilopascals) of pressure, since a uboot is not pressurized).



Oxygen:

Later in the war oxygen problems would become more acute.

The snorkel provides air to the entire ship. In order to prevent the Diesels from taking in seawater and damaging themselves by running on a vaccuum the air intake went into the engine room (not directly to the engines). This supplied the Diesels and crew with "fresh" air.

In earlier snorkel models if the snorkel went underwater a valve would close, and for a short time the Diesels would actually take air from inside the ship (lowering the airpressure and burning through oxygen). Later models instituted an automatic shutoff to make this less frequent.

As Allied air cover becomes more persistent, the boat will have to spend longer periods underwater. Longer periods underwater means more time forced to use reserve oxygen.


Overall

To me it seems like focusing on CO2 would make more sense.

1. The carbon dioxide isn't as renewable as the oxygen so it dips into its reserves more quickly. After 6 hours submerged you've probably used at least 3 hours of your dedicated CO2 scrubbing, while you've used none of your reserve oxygen.

[Conversely, if your trips underwater are very long (without the snorkel for replenishing air), then oxygen demand becomes more important. If you're forced underwater for 80 consecutive hours you'll have maybe 4-6 hours of reserve oxygen left, but you'll have more than 48 hours of carbon dioxide scrubbing left.]


2. Carbon dioxide scrubbing takes place at all times, so it's more difficult to ration than oxygen. The Diesel engines, crew breathing and fire will also use up a small amount of scrubbing capacity even when surfaced (although the chemical used for scrubbing seems like it's vacuum sealed, so it would only need to be used when necessary).


3. Carbon dioxide toxicity is more of an immediate threat than oxygen depletion. Breathing oxygen from a 21% to 17% concentration will also increase carbon dioxide concentration by that same 4%. Adding more oxygen to the system will provide more breathable air, but will not reduce CO2 toxicity (and the crew breathing the new oxygen will compound the problem as they metabolize the oxygen into carbon dioxide).

(3. also implies that you could add in more oxygen via compressed air so long as you had the chemical scrubbers necessary to take care of the carbon dioxide...which means oxygen would be even less of a problem)

[h.sie]

@UltimaGecko: Thanks again for your detailed answer. To be honest: Internally, LGN1's and my model of course still has to use the CO2-system of sh3, but we added a non-renewable CO2 component to it.

From these two internal CO2 components - the non-renewable and the renewable one, we simply calculate two O2 values (for cosmetic reasons only), which are displayed/available for the player (because for me, an O2 value of 20% is much more comprehensible than a CO2 value of 1%). This is not correct, I know, since O2 and CO2 are not 100% correlated. But there is no other (simple) solution.

So if we decide to switch back from O2 to CO2, the internal model and the gameplay won't change at all. We'll only get CO2 values displayed instead of O2 values. We can make a user poll: Do you want CO2 displayed or O2? But it has no effect on gameplay.

The only thing we can change is the balance between renewable and non-renewable component.

Modelling CO2 AND O2 independently from each other would require a huge effort.

You seem to be scientist with a huge standard. But this is still a game - no simulator. My patches won't change that situation. I cannot offer an exact model of the reality, I can only offer a coarse approximation.

Please keep in mind: I don't have the SDK, so I have to program in Assembler, not in a high-level language like C++.

h.sie

[UltimaGecko]

I don't have any sort of set standard. I've been content with the way SH3 handles oxygen and carbon dioxide since I've had the game (it's obviously not an ideal or completely accurate situation, but neither are wind speeds limited to 15m/s or a lack of water surface transparency).

My conclusion is just a recommendation based on my interpretation of what I've seen (I'm sure there are volumes of useful information hidden away on a secret webpage and in the photo archives of random cities in the US, UK and Germany).

I figured I would provide as much information as I could find so that you can decide how best to use it. Some of this is in hard numbers (like the amount of oxygen in a 50 liter flask) and physical characteristics (the toxicity of carbon dioxide being linked with its partial pressure and not its concentration). All in the hopes that they'd help you decide whether displaying oxygen or carbon dioxide is more important.


Any choice is going to vary from the historical set-up somehow - even if someone went through the arduous task of implementing measurements for hydrogen, carbon dioxide and oxygen independently they'd still be inaccurate representations. I'm pretty sure most u-boats had to rely on using a system of burettes or pipettes to measure air content and not a fancy, easily legibile circular gauge like the engineer's report has. Of course, watching the engineer mix reagents and prepare a group of pipettes probably isn't as interesting as watching him shout into a voice tube.

[h.sie]

@UG: Thanks again. I'm sure LGN1 will read your post and take it into consideration for possible changes/optimisations of the model. I have not the knowledge to decide.....

h.sie

[Hitman]

@Hitman: Bouyancy is still on my (growing) todo-list, although I HATE that topic in the meantime. Bouyancy was one of the reasons for my rest from modding. I fear to touch it again too soon, because I fear to lose my interest in sh3.exe modding at all.

Your 2 options above differ from Stieblers solution (his silent-running mod). So we have to decide what's the objective before I start to program in the future (if I start).

Sinking Rates (in cm / minute) in dependency of: (Speed, Depth, Silent-Running)???
This is a 3-dimensional problem.

Also a problem are the different buoyancy models of the supermods, especially GWX and NYGM (WAC and LSH have neutral buoyancy, I think).

Sinking rates adjusted for GWX may be not fine for NYGM. So I'll have to program only for neutral buoyancy. That means: Users of NYGM and GWX first have to disable the bouyancy model of their supermod.

This all was too complex for my simple mind.

I agree, that's why I proposed a simplier solution. BUt please don't get yourself burned, simply ignore the question until you feel like giving it a go and then give a shout so we can discuss it

As for supermods: Obviously you would do a single model and then anyone who does not like it or has conflicts with his supermod would not enable it. I think that if you are able to get a good model we will be willing to revisit NYGM and GWX models to make unofficial but authorized patches that rip the buoyancy out and let your own model shine

[h.sie]

@Hitman: Ok, that's a good idea. First I'll do some little other fixes, and if my motivation is back, I return to bouyancy. h.sie

[LGN1]

UltimaGecko, thanks again for your detailed post

As h.sie has already mentioned, modelling both components is too much work and IMHO over-kill. There are more important open issues with SH3.

So what we basically have are two parameters to adjust. The amount of renewable and non-renewable O2/Alkalipatronen. As long as the amount of non-renewable O2 is not extremely high, the impact of the value of renewable O2 should have a much larger influence on the game-play than the non-renewable amount (because usually in SH3 you dive a lot, but not really for extended periods, except during a Gibraltar passing, during the Norway campaign, and late in the war before the snorkel is available).

When h.sie and I started to discuss this topic, I proposed something of 5h of renewable O2. However, I think this might be a too small value. The current version is considerably closer to the stock SH3 situation. What I really would like to avoid is that the value for renewable O2 is too small and h.sie gets tons of user request to change the values. That's why I think a more conservative (i.e., closer to stock SH3) approach is better at the moment.

I guess the best thing at the moment is to gather more experience with the current setting, see what players report, and then adjust the parameters accordingly (maybe via a poll as h.sie has proposed).

Cheers, LGN1

EDIT: I'm just preparing for departing on a patrol to the US coast in mid-1943 and want to gather some experience with the current setting (however, it might take quite some time before I really depart because of other duties in port, i.e., the SH3 workshop ).

[makman94]

Originally Posted by Hitman

....

@H.Sie, did you finally shelve the idea of implementing real buiyancy, i.e. the UBoat changes depth when moving too slow for the dive planes to produce an effect?

Until now it was a compromise, either positive (GWX) or negative (NYGM) and never the two things depending on the situation as the real submarines do.....

the point is that i never understood this positive buoyancy at gwx that i have read many times at posts.
in NYGM ...yes, i see my sub going down if it is stopped or running very slowly....is this what you mean when you say 'negative buoyancy' ?
BUT in gwx ...my subs DON'T CHANGE depth (not even a meter up ...not even a meter down)...either if i am stopped or running very slowly . so ,my question is ''what do you mean when you say positive buoyancy' ? does this has to do with the sub's depth (tents to go up) or it is something else ?

ps: at your gwx installation : does your subs goes up when stopped or running slowly (i am asking in case that i have corrupted files...although it is a clean gwx3-no mods-install)

[Wolfstriked]

Makman I notice the down buoyancy in NYGM but see neither up or down in GWX like you.I will be honest and say I thought that was so kool in NYGM when I first saw it but then started wondering is it realistic.There is a guy in command room standing in front of a multitude of valves(Diving officer?).Isn't his job to play with all the valves to fill drain the main ballast and the trim fore/aft tanks?So when your at 0knots he would add/subtract to keep the bouyancy and hence depth the captain asked for.

[makman94]

Originally Posted by Wolfstriked

Makman I notice the down buoyancy in NYGM but see neither up or down in GWX like you.I will be honest and say I thought that was so kool in NYGM when I first saw it but then started wondering is it realistic.There is a guy in command room standing in front of a multitude of valves(Diving officer?).Isn't his job to play with all the valves to fill drain the main ballast and the trim fore/aft tanks?So when your at 0knots he would add/subtract to keep the bouyancy and hence depth the captain asked for.

i don't know Wolfstriked if it is realistic or not but i think i have read at forum that it was impossible ,when sub was stopped, to keep it at periscope depth,,,needed a less speed to achieve this.but i don't have the knowledge to answer this question (sure you will get an answer at this by someone here)

[Wolfstriked]

I am very new to this sub physics topic so I may be blowing smoke here.Could it be that at periscope depth the turbulence from the waves above cause the boat to rise and fall.By using speed and the wash over the hydroplanes you force more control?Could be why they say that subs have a hard time maintaining periscope depth only.

The compressed air could use some realism tweaks.Right now it only reduces air when you blow ballast(how much a uboat blew per this instruction would be helpful).But any time the boat changes to higher depth I would imagine compressed air is lost.There is no change in game though.


EDIT....If the sub has 3m af water above the conning tower and the water above swells and drops down then all of a sudden the boat has 2m or even 1 meter above the conning tower does this change the bouyancy of the boat.If I am understanding sub stuff right then less water above the boat means you change the displacement of the boat and now it wants to (surface?/dive?) due to that.I do not think thats modeled in SH so in effect the positive/negative buoyancy trim is realistic in that it forces one to have to use speed to compensate.....