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

[h.sie]

@Leitender, LGN1, Hitman et. al.

I have checked it in the debugger. we cannot set the relative speed (see Uboats .cfg file) to a value greater than 1.0. It must be in the interval [-1.0 ... +1.0]. Even if I force this value to be 1.1 with the debugger, the speed does not rise. there seem to be delimiters somewhere in the code.

(ok, this could be changed with coding, but I fear to break other things. so I won't touch this behaviour).

that means:

1) if we want to model an additional machine state (3x Wahnsinnige = diesel overloading) this must be connected with the relative speed value of 1.0.

2) if we want 10% extra speed at this overload state, we must enlarge the max. speed of the Uboat in the .sim file about 10%.


For the realisation/activation of this new "overload" state we have 2 ways:

A) The easy solution. Assign "diesel overloading" to FLANK speed on the machine telegraph. Relative speed value: 1.0. Drawback: We lose one regular speed level in the machine telegraph. Until this day I thought, FLANK is equal to diesel overloading. I have a private sound mod. When I choose FLANK, I hear the dramatic sound "Äußerste Kraft voraus, alles was drin ist, LI".

B) The hard solution: Add GUI elements, a new command and the necessary code infrastructure for a brandnew machine state "overload" with speed value 1.0. Reduce the relative speed values for the 5 speed levels in the Uboot.cfg file to values below 1.0. We don't lose a speed level, but I fear a huge amount of programming work for this extra machine state. Worth the effort?? In the next time I will only have spare free time. Unfortunately, this is my hobby, not my profession.

[LGN1]

Hi,

I did a quick test. I set the flank-speed value in the u-boat's cfg file to 0.5. As expected, I got only half the max. speed from the sim file when I ordered flank speed in-game. However, when I switched from the telegraph to the knotmeter and ordered 20 knots, I got the max. speed of approx. 18 knots.

So, I don't think it's necessary to create a new speed setting. It's enough to fix the max. speed value in the sim file to the max. possible speed and adjust the flank value in the cfg file to give the max. continuous speed. If the player wants to 'over-power' the diesel, he uses the knotmeter to order the max. speed. The only drawback is that there is no warning about the increased damage risk (this might be cured by changing the CE's answer when ordering a specific speed value, e.g., 20knots).

But still we need to know the max. speed with and without 'over-power'.

Cheers, LGN1

PS: Maybe it's a good idea to have just two failure probabilities. A very small one for speeds up to the flank setting and a larger one for speeds exceeding the flank setting.

[Leitender]

as long as we don't know the max. speed with the different settings, it's pointless to discuss the power settings.

Without the speed data we can always argue that the flank setting/speed is always safe.

I´m sorry, but I can´t follow. As we don´t have a gear box, speed is constantly depending on turning rounds (leaving slippage constant). If we have the turning rounds, we have have the corresponding speed. Just tested it:

Type IXC, GWX 3.0 Single mission, normal weather conditions

load_______rpm_____ speed ___ratio _______engine properties______ratio
_________________________rpm/rpm_max ____(uboat9c.cfg) ___ rpm/speed

slow______ 150 ______ 5 _____ 0,3 _________ 0,28 ________________30

onethird___ 290 ______10 _____0,58 ________ 0,57 ________________29

standard___400 _______14 _____0,8 __________0,8 _______________ 29

full________470 _____~16-17 __0,94________ 0,94_______________ 28,5

flank_____ ~500_____ ~17-18 ___1___________ 1_________________ 28,5

Result:

1. ~ 29 rpm correspond to 1 knot (correspondence confirmed with another test)

2. engine properties within uboatxx.cfg correspond to rpm and thus to speed

According to the increase shown in the former real u-boat data chart (remember 470 rpm max, 480 rpm 2hrs, 490 rpm 1/2 hr) the corresponding engine properties are 1,02 resp. 1,03. which is an acceleration of only 0.3 to 0.5 knots #

Furthermore i fear that your suggestion of adjusting some percentage won´t work, because the charging parameter "Speedboost", found in basic.cfg should work for all load steps, not only for flank (like a supercharger does in real life!) Please be aware that load and speed are different form each other.

h.sie

To answer your meantime post: Adjusting speed is also possible via uboatxx.sim. via eng_speed AND eng_power!

(But I have some testing charts from long ago in mind, where fuel consumption was measured, dependant on if a supercharger was used or not. In the case of used supercharger, top speed also arised, IIRC. was that wrong?)

Next test: backstandard=1.03 Result: ~18kn. So it works, but side effect: Wrong indicator position. Not good.

Again LGN1

Possibly the best idea! Take my parameter for choosing the correct overload speed, and maybe we got it!

[LGN1]

Hi Leitender,

just a quick comment about the super-charger. GWX has deactivated the super-charger boost because all u-boats were equipped with it before the war started (according to the GWX manual). So we have to take care that the mod works with and without super-charger.

And we need to know whether the max. speed quoted in the literature is max. continuous speed or max speed.

Cheers, LGN1

[LGN1]

Hi Leitender,

as I understand you, the three quantities speed, RPM, and 'engine properties' all scale linearly. We know that max. continuous RPM is 470 and max. RPM 490, i.e., RPM is increased by 20/470 = 4%. Therefore, speed also increases by 4%. So, what is missing is either the max. continuous speed or max. speed.

For instance, if we know the max. continuous speed v_max,cont we can calculate the max. speed: v_max = v_max,cont*1.04. We can enter this value with RPM 490 in the sim file and set the Flank value in the cfg file to 0.96 (GWX value for Full is 0.94, i.e., pretty close to 0.96).

In H.Sie's mod it's enough to modify the speed dependence, i.e., scaling the probability with a speed factor that is close to 0 for values <=0.96 and then grows fast to 1 for a value of 1, e.g., speed^64 or speed^128

Since the value for 'Full' is actually quite close to 0.96, players who don't like to change their sim and cfg files can just interpret 'Flank' as 'over-power' and 'Full' as max. continuous speed.

To summarize, I think it's enough if H.Sie modifies the speed dependence a bit and increases the probability so that T50 at speed=1 is reduced (maybe one hour ). Players who want to have a 'safe' Flank setting must adjust their cfg files.

Cheers, LGN1

PS: There is one issue I don't understand yet. If speed is determined only by RPM, where is the power difference between the two diesel configurations going? In the link I've posted, both diesels run at 490 RPM but the power output is quite different. Did they use different ship's screws?

[Leitender]

GWX has deactivated the super-charger boost because all u-boats were equipped with it before the war started (according to the GWX manual).

I saw that too. Maybe due to h.sie´s finding above about the limitation of "engine properties" to "1", i.e. 100%, but not sure.

we need to know whether the max. speed quoted in the literature is max. continuous speed or max speed.

It is common to name that max. speed under which secure operation is guaranteed.

the three quantities speed, RPM, and 'engine properties' all scale linearly.

It seems to be so, though I was amazed, too. I thought that machine telegraph settings should correspond to load, not to rpm. But obviously...

So, what is missing is either the max. continuous speed or max. speed.

i´m sure that "flank" is max continuous speed at 470 rpm. In literature there is named a value of 19.2 Knots or similar for type IX C (with supercharger, I think)

The reason why in my test the max speed was a little bit lower could be crew efficiency lower than 100% or other less important parameters.

(Don´t ask me why i came to 2% and 3% speed increase. This should be 2% and 4%)

If h.sie is right an we can´t "accelerate" the engines beyond 100%, we have to set max speed to 100%, which is max continuous speed * 1,04. That´s correct. But max continuous speed should set to flank.

If we set max speed to 19.2 Knots *1,04 = 20 Knots and then: aheadflank=0.96, we shoud get 19.2 Knots under normal conditions.

we also can adjust the .sim-setting of eng_rpm from 490 to 470. But force is set to 4400 bhp (!) instead of about 2500 hp in reality. With this value we also can change speed settings. Set realistic value of 2500 into .sim led to a lower max. speed. Set to 44.000 bhp (10x) led to higher max speed of 21 Knots, but much lower than expected (?).

If speed is determined only by RPM, where is the power difference between the two diesel configurations going?

Speed ist not determined by RPM, but is a linear function of rpm. As mentioned above, I did a second test with real life values insert into .sim and this led to a factor of continuosly 31 rpm per Knot instead of 29 rpm per Knot. Not too much, but significant.

To answer your thought, I think that when the starboard engine is in battery loading status, this factor is reduced to somewhat of 50-60%.

At the end of the day i wonder if h.sie is able to program his damage probability dependant on this speed/rpm-factor. I assume that the functionality up to now depends on machine telgraph status, not on Knotmeter.

[h.sie]

I removed the delimiters in the code so that I could set reative speed = 1.1 = 110%.

Results:

Speed goes up from 17kt to 19kt. Fine.

But: RPM sank!!

Setting rel. speed = 2.0 lead to a speed of > 30knots and RPM=0 !!!)

It seems that RPM in game is for eye-candy only - for the propellers and gauges, it seems to have no physical meaning.

Unknown so far: Other weird side-effects.

This shows that with removing the (-1/+1) delimiters I have damaged an closed system of physical formulae. -> Better not touch it.

Only solution:

Max. relative speed = 1.0 must be interpreted as "overload".

Thus, in the next version of the Mod, rel. speed = 1.0 will be connected with an even higher damage probability, which results in T50% = 1h.

But all lower speeds (continuous speed levels) will have a extremely low (or zero) damage probability (provided both engines are undamaged).

The prize for the diesel damage mod is that we lose a slot for a continuous speed setting.

By the way: I, personally, don't lose that slot, since I always interpreted FLANK as machine overload.

As LGN1 wrote: People who use this mod and want FLANK to be a save speed setting, must assign it a relative speed value of e.g. 0.95. The resulting speed reduction could be compensated by increasing the max. speed in the .sim file accordingly. Those people have to activate the overload speed setting in the knots dial.

[Leitender]

Hi h.sie

thank you for your effort. With max overload speed = 1.0, reduced telgraph settings and LGN1´s "extra overload order" by setting the Knotmeter to its max value we should ride the boat. But is it possible to simulate damage at a relative speed value > 0.96?

[h.sie]

@Leitender:

In the current mod version, the damage probability per second is calculated as follows:

p0 = 1e-4 * RelativeSpeed^32

(provided both engines are undamaged). Thus, every speed setting has a damage probability, but due to the high exponent of 32 the probability drastically sinks for values below 1. Example:

For FLANK: p0 = 1e-4 * 1^32 = 1e-4 * 1
For FULL: p0 = 1e-4 * 0.94^32 = 1e-4 * 0.14
For STD: p0 = 1e-4 * 0.8^32 = 1e-4 * 0.0008


In the next version of the mod, I will use a higher exponent than 32, so that 1.0 will result in a high p0 and 0.95 will result in a very low p0.

[Sailor Steve]

If you make this work I can supply an extra voice. In SH2 it was possible to have the engines break down. It could happen any time at random, but was much more likely at Flank speed. If the engines were run at that speed for too long and trouble was eminent you would get a warning: "Engines are overheating, sir!" It's not the same as a warning about potential dangers, but it's the best I have on file. It would be easy for me to incorporate it into the Flank speed voice file.

[Leitender]

h.sie
I didn´t know that you programmed the damage speed-dependant. When i read your description of the mod i thought you made it telegraph-state- dependant. That´s why i asked.

But with this information overloading the engines via Knotmeter should work. That´s fine

[h.sie]

@SailorSteve: Would be great. Only problem: I have no experience to program / trigger new sound events. But maybe Stiebler knows (maybe he did that for his "smoke on the horizon" mod).

[h.sie]

@Leitender:

Yes, I use the telegraph setting because it is much easier and common for all Uboat types. I assume the Uboats absolute speed to be proportional to the telegraph setting.

This is true, as long as

- enough crew in the compartment
- fuel not empty
- engines undamaged.

I of course consider these exceptions in the code of my mod.

[asanovic7]

tested it.. didnt get the ai subs attack, better said didn't get the last order (attack).. individual atack recommended.. although, bdu said pack is comming and all, and then silence, and after 2 hrs again requested contact updates.. at 5 am the order from above individual attack recommended.. is that normal?

can someone tell me how does it look? submerged subs (wild wish) or?

can this stable version be used to play campaign? or it's still a test?

happy hollidays to all and cheers!

[h.sie]

@asanovic: This can happen from time to time. Maybe the weather was too bad for the wolfpack subs to come in time to the interception point.

Try again. there is a lot of randomness.

Important: If, after your 1st contact report, you get the order to shadow the convoy, you installed the mod correctly.

V16A3 is stable and can be used for camaign