pressurization in the 747-400

[michal 44]

quoting RR: The only thing left to install is pressurization. Yes- I'm dreading that one because of the mathematics involved... but that's all that's leftIf one thinks of a highly 'professional' touch in the area of pressurization one thing comes to mind which so far has been neglected by practically all simulations - proper rate of pressure change. For example pressing a dump switch would result in some controlled outflow of the air from the cabin - the valves have but finite dimensions. This sort of action typically results in an almost instanteonous (and significant) drop of cabin pressure in the current generation of FS simulators.Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2

[lefteris 70]

Michael J.-why do you think we're dreading it? :-) We always strive for realism - if it can't be done properly, it's best left out. (This one, though, I think we're doing ;-)).

[michal 44]

> (This one, though, I think we're doing ;-)).:-beerchugMichael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2

[Guest mc.kar]

Actually when I read RR posting about sorting out the math...Got me thinking. You guys at PMDG must have sorted out all the math when doing the NG. It has a cabin altitude gauge on the OHP...Or what? ;-IAnyway...good luck dealing with the MSSA (as opposed to ISA!) I have discovered that MS hasn't simulated the atmosohere that well.Anyway...there are some freeware cabin altitude gauges out there. I have made one that I'm using for the RFP747 and it works reasonably well - mathwise. I know that 747-400 needs to display the cabin altitude (and related info) on the ECS synoptic page so engineering this solution will off course be a task in itself. However there are XML scripts available whith all the math functions related to calculating cabin alt, cabin rate and diff pressure and using it. I used that since I got gray hair trying to figure out the math behind the MSSA.Just a suggetionAnd good luck with the 747-400 we are all looking forward to it.Best regardsMichael ChristiansenKarup, Denmark(mich.chr@mail.dk)

[kevinh 409]

Mathematics of a pressurisation system are quite straightforward if you've got the rest of the pneumatics/air-conditioning systems right.Cabin pressure rate is proportional to the net airflow into the cabin. Cabin altitude is then computed from cabin pressure. If you always work in terms of pressure and airflow it's much easier.Another tip, compute everything backwards from auto system demanded values (don't even try to simulate the real way the outflow valves are controlled). In other words, position the valves to give you the outflow you need to get the pressure rate demanded. This results in a stable, solid yet realistic simulation.In my experience (real world flight simulation) the pneumatics system is much harder to get right.If you need any help guys... anything to get the Queen of the Skies released earlier :D

[michal 44]

>Cabin pressure rate is proportional to the net airflow into>the cabin. I am not convinced things are so simple if one wants to simulate all the modes of pressurization correctly. This "airflow" you mention above must be a complicated function of pressure differential, altitude, temperature. If I try to dump pressure at say 35000 ft the resulting flow will be obviously different then at 15000. To model this correctly and account for constantly varying pressure differential one would have to integrate to get the cabin pressure right. And then of course temperature must be dropping like crazy inside the cabin which in turn interacts with air conditioning ... You get me on this programming team and I can complicate things with no end in sight .. :-lolMichael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2

[lefteris 70]

... You get me on this>programming team and I can complicate things with no end in>sight .. :-lolOUT OF THE QUESTION. We'd be spending our entire workday arguing about bitmaps ;-) ;-).

[kevinh 409]

Like I said it's easy. You guys are looking for difficulty. For the purposes of simulation, even on Level D sims, airflow out of the cabin can be aproximated to:K * pressure differential * outflow areaThe value of K is arrived at by trial and error. Outflow area is a function of outflow valve position and cabin leakage area. Leakage is determined by adjustment with packs off to get a reasonable leak climb rate (around 300-500 feet per min).Airflow into the cabin can be as simple asX * number of packs runningIf you're modelling pack airflow then it's even easier, just add them all up.If you want to further correct for pressure and temperature (not sure it's necessary in a PC game) then treat the above as "normalised" flows and apply corrections based on the ratio of ambient pressure to sea level ISA pressure and square root of the ratio of ambient temp to sea level ISA temp.The different modes are taken care of by the controller simulation, which really just programs demanded cabin altitude as a function of aircraft altitude and rate of climb, with safety limits added. The cabin air behaves the same regardless of what mode you set.If you already know all this then say so.Seriously, if you want help or advice from an experienced simulation professional just ask. I'd be glad to be of help. Remember, a pilot will know how an aircraft system works, but usually not the best way to model it. Send me a contact email address via the forum.Kevin

[michal 44]

>K * pressure differential * outflow area. and probably we can assume (for the purpose of this simulation) that outflow area is another constant since who would know how to simulate the actual position of the valve. Then the above further collapses to const*press.diff. And the constant can be derived from what you describe as the reasonable leak climb rate (~500 fpm). So dumping the pressure would probably result in the same amount of leak rate or can we assume it would happen faster (if valve is fully open for this case)? Love this stuff ...Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2

[kevinh 409]

OK, what makes you an expert Michael? Most of your comments to my inputs are (a) unnecessarily sarcastic, (:( unhelpful and © missing the point. Or do you just enjoy stirring up arguments?I sense from your comments that you don't understand the concept of modelling a system at all. All you can see are complex equations to re-calculate each parameter with a series of special cases to cover each possibility, rather than the iterative approach used in "real" flight simulation.I'm not going to describe simulating the outflow valves just so you can take the p!ss but take it from me it's straightforward.And no, outflow area is not a constant. Re-read my earlier post.If PMDG want to respond to my suggestions please let them.Why are you so concerned about dumping pressure by the way? If you model pressurisation as I suggest all such possiblities are automatically catered for... no special cases needed.

[michal 44]

>OK, what makes you an expert Michael? Most of your comments>to my inputs are (a) unnecessarily sarcastic, (:( unhelpful>and © missing the point. Or do you just enjoy stirring up>arguments?Perhaps rather than looking for conspiracy theories just read what I wrote and stop looking for hidden meanings. My comments where neither sarcatisc nor missing the point nor something else. I was just discussing plain physics involved here. I focused on the outflow area as one thing I knew nothing about. I hoped you could lead me to explanation how this paramter is modeled (in case it can't be assumed constant). And by the way, you are not the only one here who is/was simulation 'specialist'. So am (and I do mdel 'systems') but I admit not aircraft hardware/systems. And the reason I 'focused' on dumping pressure is that such button is to be found on many (most?) pressurized aircraft and pressing such button results in 'drastic' action and drastic actions are usually hardest to simulate. But if you can't stand technical discussion in an open forum without need for some personal adjectives - so bid.Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2

[kevinh 409]

What simulations have you worked on? I assume you mean that your experience is with FS add-ons. I'd be interested to know if there were any in my MSFS collection. My experience is with commercial flight training simulators.You misquoted parts of my posts out of context, apparently just to make fun. You are so sure it must be complicated that you won't consider a simple alternative. Of course an aircraft pressurisation system is very complex, but the simulated version need not be a direct emulation. The response from PMDG seemed to indicate they know you and expected pointless lengthy debate. You even joked about complicating things endlessly given the chance. My main aim here is a genuine attempt to help PMDG reduce their development time and get the 747 out sooner.Simulation is not all physics and maths models do take unscientific shortcuts. Whenever I see the claim "accurate physics model" I get very sceptical. If you try and calculate pressurisation parameters mathematically you end up with the difficult equations you mentioned. If you model the system dynamically, rather than statically, you don't have to solve these equations. The model acts as a continuous numerical solution. You must realise this model operates iteratively. It will not compute an instant answer, but integrates the solution over time. Is this making sense yet?Basically imagine the cabin is a large balloon with small hole in it. You have to pump air into it to keep it inflated. In reality the amount you pump into the cabin is not controlled, rather the size of the hole letting the air out, which is what outflow valves do. On a 747 these valves are massive. If there is more air going in than escaping, the cabin "inflates" (pressure increases, cabin altitude decreases) and vice versa.The outflow valve (in fact there are always at least two but they usually move together) is modelled as a variable leak area. It is controlled to give the required net airflow (airflow from packs minus total leakage) to achieve the cabin pressure rate demanded by the pressure controller. You can then integrate to get cabin pressure and calculate the cabin altitude for indication. This is not much like how a real pressure controller drives these valves, but it looks correct and is much easier to get stable.The various pressurisation modes are simulated via the demanded pressure rate so the cabin maths model remains the same. In manual mode the outflow valves are driven by the user and the cabin is allowed to respond.In reality if there is a sudden large pressure loss, no cold outside air gets in. The conditioned air in the cabin escapes. So there is no sudden temperature change. If you want to model the effect of cabin temperature on pressurisation you can, but it's very small. Just don't try and solve the maths :)The whole point is you program the model and let the PC compute the solution, you don't solve the equations yourself. That's as technical as it gets.By the way, I'm not aware of any modern commercial airliner with a dump switch as you describe. You can manually open or close the outflow valves of course, though only gradually. There is no need for the crew to be able to initiate such a decompression, other than to manually depressurise after landing if the automatic system fails. Some aircraft have a ram air inlet. This is not a dump valve and will not cause a loss of pressure if opened when pressurised (though this is not recommended). I agree that most FS addons either set cabin altitude as a function of aircraft altitude, or straight to ambient with no intermediate modelling. There are exceptions, notably some freeware panels.Hope this helpsKevin

[teropa 4]

Michael,Have you tried this with LDS ? ;)Tero

[Guest rcritz]

>By the way, I'm not aware of any modern commercial airliner>with a dump switch as you describe. You can manually open or>close the outflow valves of course, though only gradually. >There is no need for the crew to be able to initiate such a>decompression, other than to manually depressurise after>landing if the automatic system fails. Some aircraft have a>ram air inlet. This is not a dump valve and will not cause a>loss of pressure if opened when pressurised (though this is>not recommended). I agree that most FS addons either set>cabin altitude as a function of aircraft altitude, or straight>to ambient with no intermediate modelling. There are>exceptions, notably some freeware panels.Kevin,I was 100% with you until this point. I don't fly an airliner but a FAR 25 (transport category) bizjet. It absolutely has a dump switch whose purpose is to command both dump valves wide open, resulting in a cabin climb rate of about 4000-5000fpm. They are hardware limited to close when the cabin altitude exceeds 13500 so if you still have ACU air, your cabin will oscillate in the vicinity of 13500. Without ACU air, the cabin will continue to climb above 13500 at the leak rate (about 800-1000 fpm). But our checklists absolutely call for the dump switch, particularly for getting smoke out of the interior environment. Without spending much time researching FAR 25, I'd be surprised to find other a/c certificated under that part without similar capabilities and needs.But, as I said, for the purposes of modelling the system, I agree with you 100%.Oh, and our primary use for this switch is to absolutely insure that the cabin differential is as close to 0 as possible to minimize the "bump" that happens when we open the airstair. Same reason we always try to initiate the start sequence as the door is closing with pax on board. Start means the ACUs are off so there is no inflow and no "bump" as the door closes against its seals.cz

[kevinh 409]

My experience is with commercial airline flight simulators (most Boeings, Airbus, MD80, Fokker). None I have encountered have such a system you described, though in all cases they have an auto pressurisation system which ensures minimal pdiff on touchdown and dumped pressure soon after (at a rate of about 500 fpm). If this doesn't function, the procedure would be to switch the system to manual and open the outflow valves fully.To protect against excessive differential pressure they usually have pressure relief valves which open automatically in a schedule to limit pdiff to maximum values. Sometimes a pack is auto-tripped off to assist this. There are also inward relief valves to protect against any negative pdiff.Is your aircraft's dump switch part of normal procedure or an abnormal procedure in the event of auto system failure? If the former, I guess the auto controller must be fairly simple (less to go wrong and maintain).As with your bizjet, most airliner auto pressure controllers will drive outflow valves fully closed when cabin altitude reaches the level you suggest, but that is the opposite of dumping the pressure, and I doubt your passengers would appreciate you operating the dump switch at altitude. Under what conditions would you operate the switch in the air?