Sim9b feedback

[Dr.Sid]

Check out the new version:

http://www.commanders-academy.com/comsubsim

Post feedback related to this version in this thread.

[To be]

The sub really tends to gyrate along the long axis. This is most visible when diving or surfacing, or trying to use the depth autopilot at all (it doesn't work, but it shows the problem) The rudder on the subs react instantaneously to commands, while the planes don't. Buoyancy seems a bit low - after blowing the tanks and coming up at a steap angle the sub tilts nearly 40 degrees downward after hitting the water. Perhaps the 'capacity' of the ballast tanks should be changed? That is really just an impression.

[Dr.Sid]

Don't understand .. you think my subs rotates too much or too little ? Rudder and rear planes should react more or less same, and they do for me. Can you describe the situation a bit more (sub, speed, dept). Or do you mean the rudder itself moves instantly into maximum deflection ?

Buoyancy is derived from wikipedia values (for Los Angeles, Akila in fact has same weight/volume at the moment). It should be more or less correct. However the dynamic stability might be a bit low. It's the tendency of the sub to turn in the direction it's moving.

[To be]

Sorry for not being clearer:

#1 The submarine is hard to stabilize along its length, and the depth control autopilot fails totally.

#2 Try an 'emergency surface'. Go to the bottom, order flank bell. Then press S to blow tanks, order all planes full up. The sub emerges from the water as expected, with a high up angle, and then splashes down. However, after it hits the water, the nose pitches way over 40 or 50 degrees downward, and the tail of the sub comes right out of the water. This is very similar to how subs surfaced tail first in DW.

#3 The control surfaces move instantly to wherever is commanded (mostly just something to remember in the future.

[Dr.Sid]

I see ..

#1 .. I know about some problems with Akula depth autopilot, which can get quite mad, because front planes affects pitch a lot. I'm working on fix. However the situation was quite rare. I still demand more detail :biggrin:

#2 .. understand .. I will postpone this for the moment, since there will be same changes witch would affect this. I have to take turbulences into account on surface, witch will increase drag on surface, and maybe even surface tension acts here. I mean now the sub dives too deep after emergency surface. Sail should not get submerged from what I've seen in videos. In such case the vertical speed will not get that fast and rear planes wont turn the ship nose down. As I said .. let's tune this later.

#3 .. understood .. known problem. Thanks for report anyway.

[Hitman]

Oh, I had posted in the other thread before, may be you thought I was referring to the older version :mad:

Never mind....

I don't see any implementation of derivation yet, is that correct or is it simply too subtle and I don't notice it? :confused:

[To be]

I am doing some testing with the autopilots, and I can't seem to reproduce my last problem. Any reason the depth autopilot is limited to 1500 ft?

Additionally the speed hold autopilot seems to not hold the speed request, but some other speed, depending where you are. Additionally, pressing D when using the tug = dive dive dive!

[XabbaRus]

I have noticed that ship speed and ordered speed can be about 2 knots less that what is ordered..

[To be]

I found it varied from 1 knot less to 3 knots more as I moved around the map. It might depend on location, or depth... I don't know.

[Bubblehead Nuke]

I see ..

I mean now the sub dives too deep after emergency surface. Sail should not get submerged from what I've seen in videos. In such case the vertical speed will not get that fast and rear planes wont turn the ship nose down. As I said .. let's tune this later.



It most definitely go back underwater if you start it deep enough. Now if you are starting an emergency surface heavy (due to flooding) then you will NOT surge out of the water but look like the oil diesel boats did when they drive to the surface. The bow will come out and then the boat will tip forward pretty much level depending on trim.

If light at normal trim, the boat will surge out of the water, broach over and then go down at essentially a level attitude.The boat will then sink out, recover, and then come up pretty much level and on the surface. The amount of sinkout is a function of intertia.

But I agree, tune this later. This is fine tune stuff.

[Dr.Sid]

Max ordered depth is now limited, and 1500 just seem nice number to me. There is no more reason to that.
Speed order now works in very simple way. Frontal drag coefficient of hull alone is taken into account, required force is computed for it, and such force is used as thrust. Nothing more.
So on surface the sub will go faster, because now on the surface there is less frag because of less frontal area.
Submerged the resulting speed is somewhat slower, because of planes and sail drag.

Tug can dive so you can pretend you have sink it ;-) Actually I like it sitting on the seabed. It looks so peaceful.

[Red Ocktober]

So on surface the sub will go faster, because now on the surface there is less frag because of less frontal area.
Submerged the resulting speed is somewhat slower, because of planes and sail drag.

you're aware that this is only true for certain hull forms... that is partially why the Skipjack form was adopted over the previous conventional hull designs... it was capable of significantly higher submerged speeds than the conventional form the Nautilus/Skate subs employed...

this is one of the problems of attempting to employ realistic physics modelling approaches in a sim... you've gotta make sure it's based on correct hydrodynamic models, and that the results be derived from each is based on the correct data...

in this case, one size doesn't fit all...

i suggest you abandon this approach (don't kill me :biggrin: ) and go for a code (faked) physics implementation... explicitly plug in the different performance data to meet the already known results...

start with coding a simple bouyant box... then add code to adjust its bouyancy... then add code to propel it... then add code to take into account the effects of dive planes and speed...

i've gone both routes... this one is easier unless you have the data for the hull form you want to use ready to be plugged into your physics sim...

(do you have a Blitz3D license? )

--Mike

[Dr.Sid]

I do it more or less like you say. I don't take exact hull shape into account. I use ellipsoid now (ie. sphere whose width, height and depth are not the same).
But you simple have to scale all effects based on how much the part is submerged. It means less drag on surface. And there is no problem with that.
The problem is I don't take any kind of possible turbulence into account, and on surface I have to. I will simply add another drag parameter which will be used on the surface and that should solve the problem.

[XabbaRus]

This isn't feedback as such but relates to the physics. When we get to sink stuff eventually are the ships going to sink in a different manner according to where they were hit, sea state etc? Would be interesting to see a ship or small boat capsize first then sink..

[Dr.Sid]

The physics can handle it already. We just have to define enough inner compartmens, then flood those damaged and leave the rest to the physics engine. It's in principle same as tanks in sub.

I'd like to define all inner 'rooms', have them connected correctly, define the 'holes' where the water will go in, allow pumps and blowers to be placed into them, and simulate all this (in basic terms I mean).

MBT tanks will be just special case. They flood in perfectly same way as lets say command room, except you usually flood MBT on purpose.

So we will be able to simulate failed torpedo door, dive with upper hatch open, or just waves coming over it, and of course any battle damage.

There are areas where to simulate stuff correctly is impossible (in real time), and you must fake, but this is not the case.

[Dr.Sid]

Btw. I've just checked the torpedo speed. At 50kts it should pass 688 which has 100m in about 4 seconds (50 kts is more or less 25 m/s). And it does.

[XabbaRus]

OK that's cool. For subs I won't break up the model as I don't see a sub splitting in two just a bloody big hole in it when hit. I was thinking we can handle that with alpha maps and some sort of dented damaged model.

As for ships I'll have to start looking at that.

[Dr.Sid]

No need .. I would leave this to some 'second' phase. Right now all I want is few good subs (lets say Akula and 688 since they are known from DW), but let's make them good. I think the shape is good, polycount is good, even on models I have. We need all moving parts on them, some details, tune textures. I think you don't have to make all the models, but we should define some standard of detail a quality.

[Molon Labe]

OK that's cool. For subs I won't break up the model as I don't see a sub splitting in two just a bloody big hole in it when hit. I was thinking we can handle that with alpha maps and some sort of dented damaged model.

As for ships I'll have to start looking at that.

That and no one will ever see it, except maybe as separate sonar returns. For a skimmer, visuals are useful for BDA.

[goldorak]

No need .. I would leave this to some 'second' phase. Right now all I want is few good subs (lets say Akula and 688 since they are known from DW), but let's make them good. I think the shape is good, polycount is good, even on models I have. We need all moving parts on them, some details, tune textures. I think you don't have to make all the models, but we should define some standard of detail a quality.

Lets give a little love to some surface vessel. :biggrin: (arleigh burke or perry frigate, preferably the former, but just as well the later).
You know, to give those bubbleheads something to worry about.

[XabbaRus]

Hey I'm coming to the home stretch with the Oscar II. It's level of detail is more than anything I have made before. Poly count is still good too...Then I'll fix the 688 and give you an Akula with opening doors.

[Zachstar]

So phase one is 688 and Akula-1 eh? Sounds like one of those old Apache vs Havoc games.

[Dr.Sid]

More or less. I mean there are subs we all know from DW. While my plan for sure is not pure copying, DW is where the reals business starts and we must get there first.

[Zachstar]

Just getting to Sub Command level will be a success!

I suggest phase 2 include the Kilo (Kilo basic likely the one from China) To experiment with a number of torpedo types and hopefully far better battery simulation.

[Deamon]

The problem is I don't take any kind of possible turbulence into account, and on surface I have to. I will simply add another drag parameter which will be used on the surface and that should solve the problem.
What about reducing thrust instead ?

Ever considered how close the screw on a modern sub is to the surface, when it runs on the surface ? It basically scratches the surface all the time and drags in a lot of air around the screw and this kills the thrust, imo.

[Theta Sigma]

I'm not sure what the issue is on my end, but on Vista 32-bit + SP1, Sim9 and Sim9b both crash. :(

[Dr.Sid]

What hardware do you have ?

[Deamon]

I had quite a bit of fun trying to stear the Akula at PD. But I found it annoying that the pitch indicator doesn't goes into the floating point range. I couldn't spot the tendency early enough and kept bouncing up and down.

Any chance to fix this ?

[Dr.Sid]

I had quite a bit of fun trying to stear the Akula at PD. But I found it annoying that the pitch indicator doesn't goes into the floating point range. I couldn't spot the tendency early enough and kept bouncing up and down.

Any chance to fix this ?

In next release there for sure will be angle with 1 decimal place, and maybe even gauge with needle (I have them basically working, I just need to configure them to display correct values in correct ranges and so).

As for the surface speed you mentioned earlier. As I have found recently, for surface ships most drag comes from making waves. It's different with submerged sub. So I will add this special component of drag as stand-alone parameter.
But yes, prop too will be less efficient near surface, it will cavitate at low speed. There is even difference between top and bottom part of the screw near surface so the boat turns slightly when main prop is loaded. I don't intend to simulate this at the moment, but it sure won't be any hard.

[Bartolomeus]

Hi

Iam tried the newest version....and i must say, nice work on this project.

My question is, it is possible to enlarge the depth in open waters?

The second question is, the type of the model files is fob. What kind of program i need to create these files? Or it is possible to export the files in 3d studio max?

Marko

[Dr.Sid]

The FOB files are mine, and the format is not yet stabilized. They are designed so they load fast and are ready to use with no additional processing.
Normally I use 3DS as source and that is the format I get data from modelers in. FOB is just geometric cache, but once it's there, 3DS are not needed. 3DS versions will be distributed with full version, if their authors will agree.

As for the water depth in areas uncovered by detail map, there is no problem, will do.

[Bartolomeus]

Sounds great! Iam looking forward to the new versions of this great project.

Marko

[DocNobby]

Hi!

This project is coming along really great.
It's already fun to play! :biggrin:


Looking forward to the next release... :rolleyes:



Greetings,

Doc Nobby

[Theta Sigma]

What hardware do you have ?

My apologies for taking this long. ;)

It crashed on 2 different Intel laptop GPUs (965/X3100 and 915). Not exactly gaming h/w, I know. However, I tried the older beta on the 915 pre-Vista upgrade (XP) and it ran fine.

[Dr.Sid]

Did you try to disable VOBs in cfg file ? It's needed on those chipset. I still can't autodetect the trouble.

[Theta Sigma]

I set it to =0 and that seems to resolve the CTD, but it's extremely sluggish. I get about 1 FPS.

[Dr.Sid]

Try to decrease the visibility range with ctrl+f1. It works on Asus EEE, so I guess it should work anywhere.

[Theta Sigma]

Nah, no dice. I reduced visibility to 12.5km and zoomed out, and the FPS was the same.

[Dr.Sid]

Hm .. can you post full spec of the notebook ? I could look if some friend of mine has something similar.

[pingjockey]

If someone wants to post the specs again I think I may be able to help test this issue.

[Theta Sigma]

Here's the lesser of the two:

http://www.docs.sony.com/release/specs/VGNFS8900_CTO_mksp.pdf

It has a Pentium M 740 Dothan @ 1.73Ghz, 2GB RAM DDR2, 120GB Scorpio and of course the Intel 915 GMA running Vista Business 32-bit.