109 advice needed (climb)

[Robo.]

Originally Posted by jg27_mc

Your one of those guys that... Well I am done with you. Take the bicycle home.

PS: (I should really be killing you instead of wasting my time here)

Not sure what you're on about here with the bicycle and stuff, but that little 'fight' of ours on ATAG was rather interesting, wasn't it?

Honestly, I hope the devs will keep improving the sim FM-wise - I believe the main FM problems are well known and documented and I am looking forward for the upcoming patches.

[jg27_mc]

Originally Posted by Robo.

Not sure what you're on about here with the bicycle and stuff, but that little 'fight' of ours on ATAG was rather interesting, wasn't it?

Honestly, I hope the devs will keep improving the sim FM-wise - I believe the main FM problems are well known and documented and I am looking forward for the upcoming patches.

It sure was (I really mean it). Hopefully this is a computer game and I live to fight another day.
Regards.

[AKA_Tenn]

so why couldn't they just let the physical model do it? give different parts of the plane weight, give the air weight, then let the physics engine do the work instead of giving the planes flying attributes absolute values... so there's no flight model at all...

since this game is really graphically intensive and hardly uses any CPU at all really... but i guess we still have a ways to go for that?

i was just thinking that because it doesn't matter weather its a machine gun or a sack of potatos if they weigh the same, they'll effect the characteristics of the airplane very similarly... so that way you wouldn't need to know the flying characteristics of the plane, just the thrust, shape and weight distribution

[ACE-OF-ACES]

Originally Posted by AKA_Tenn

so why couldn't they just let the physical model do it? give different parts of the plane weight, give the air weight, then let the physics engine do the work instead of giving the planes flying attributes absolute values... so there's no flight model at all...

A 6DOF flight model is a physics model!! Calculated on the fly (pun intended) in real time

Originally Posted by AKA_Tenn

since this game is really graphically intensive and hardly uses any CPU at all really... but i guess we still have a ways to go for that?

Nope! 6DOF flight models have been in use on PCs for some years now.. First I know of was PACIFIC AIR WAR 1942 by Microprose.. Back than it used fixed point math, but it was a 6DOF FM. As you noted, the modern CPUs have no problem what so ever running a floating point 6DOF FM. Prior to that PC flight models were, what was commonly known as TABLE BASED (read lookup) flight models that had very little physics to them (SWOTL, RB, AOTP, AOE, etc)

[AKA_Tenn]

Originally Posted by ACE-OF-ACES

A 6DOF flight model is a physics model!! Calculated on the fly (pun intended) in real time


Nope! 6DOF flight models have been in use on PCs for some years now.. First I know of was PACIFIC AIR WAR 1942 by Microprose.. Back than it used fixed point math, but it was a 6DOF FM. As you noted, the modern CPUs have no problem what so ever running a floating point 6DOF FM. Prior to that PC flight models were, what was commonly known as TABLE BASED (read lookup) flight models that had very little physics to them (SWOTL, RB, AOTP, AOE, etc)

if thats's the case why do people complain about flight models here? seems to me these guys just need to learn their plane instead of insist everything bends to their will...

[CaptainDoggles]

Guys the flight models are an approximation. Even the best gaming PC money can buy could not model the exact equations behind flight fast enough to run at playable speeds.

[AKA_Tenn]

yea see for me, i don't care about the flight model all that much as long as its not a rubber-band sim like aces high or fsx... where ur plane feels like its suspended between two rubber bands...

but it would be nice to have the actual plane itself and all its parts be modeled physically instead of just visually

guess its too unrealistic to ask someone to model just all of one materials attributes into a physics engine, never mind hundreds that would be in a airplane

one day... before i die hopefully... someone will make a general purpose physics engine that models every element on the periodic table, then we can all build our own 109s for free

[ACE-OF-ACES]

Originally Posted by CaptainDoggles

Guys the flight models are an approximation. Even the best gaming PC money can buy could not model the exact equations behind flight fast enough to run at playable speeds.

Actully PCs have been capable of calculating a 6DOF FM in real time for some time now.. As noted the first that advertised a 6DOF FM was PAW 1942 by Microprose back.. Gezzz.. I want to say 15 years ago?

[ACE-OF-ACES]

Originally Posted by AKA_Tenn

if thats's the case why do people complain about flight models here? seems to me these guys just need to learn their plane instead of insist everything bends to their will...

There is a lot of truth in that

The problem is typically two fold, one the sim pilot does not know what the value should be, and two they preformed the test wrong. Which is why it is so important to record a track file of any so called 'test'

Which is not to say there can not be an error in the FM! Just that typically it is in the sim pilot!

The FM, like any program, follows the rule of garbage in garbage out

With that said know that the 'math' of the FM is the same for all planes (subtle differences for say single vs. twin, etc). What makes a P51 fly like a P51 is when the P51 parameters are used as inputs to the FM. For a simple example say the 'math' of the FM was as follows

y = CL*x

Where

y is the output (result)
x is the input
CD is the 'drag' parameter (coefficient)

Say

x = 3
CD = 5 for a P51
CD = 5.5 for a Bf109

Than the output y would be for the given x input

y = 15 for a P51
y = 16.5 for a Bf109

Lets assume that the value of CD is not 'known' for the WWII plane we want to simulate.. In such a case you could just pick a value of CD based on other know values of similar planes (happens a lot in RL) or maybe they calculate it offline using another another program (say simulated wind tunnel) that uses some 3D model to calculate CD

In either case, a sanity check of the value you pick can be check by comparing some of the simulation results to real world results.. In this case say top speed.

Problem with most sim pilots is they don't even know what value they should obtain during a test! Let alone able to record 'all' the values that can affect your speed. For example, to test for top speed you have to be able to hold your alt within around +/-100ft.. Most sim pilots 'think' they can do this no problem.. I can tell you how many sim pilots were amazed at how much the alt varied during their flight! Which I was able to show them by using DeviceLink to record the values of altitude and speed and plot them.. And sure enough.. Most of the guys who said the plane was too fast, where actually loosing alt (shallow dive) when the obtained that max speed, and visa versa, sure enough the guys who said the plane was too slow, where actually gaining alt (shallow climb( when the obtained that max speed.

[ATAG_Doc]

That's a lot of info Ace.