Formation flying and Aerodynamic efficiency
-
-
Mav, I was under the impression (engendered by the articles available for download from this very website) that rather than each packet of air being independently calculated for each quanta of time, instead forces are applied to each control surface individually and the vectors are summed to determine the course of the aircraft, with forces varying according to a number of factors.
this I believe is what molni was getting at, rather than attempting to point out that no real air was harmed in the making of this sim.
I trust my impression was not in error? if so, I might recommend the updating of the articles on the flight model.
I hope this is not the case, as they were a most entertaining and informative read.
-
Itās already pointed in Mav-JPās articles. Tabular data is used for angular movement coefficients together with (I presume) dynamic pressure. See the article talking about Navier-Stokes equations in the title.
What I donāt understand is whether compressible or incompressible flows are used in the sim for calculating the dynamic pressure in order to calculate the pitching/yawing/rolling moment coefficients and resulting movement. Itās never pointed out anywhere in the documents.
-
What is meant by not modeling air is that there is no n-element tracking of the air medium. Computations are done with the assumption that the airās properties are irrespective to previous interactions with objects. Itās not possible to for example have an airplane change the air and that changed air to affect another airplane. There is no air medium which can take on a history or have a memory. If you want an airplane behind another airplane to feel wake turbulence then it must be done directly from plane to plane, not plane-air-plane.
-
Itās already pointed in Mav-JPās articles. Tabular data is used for angular movement coefficients together with (I presume) dynamic pressure. See the article talking about Navier-Stokes equations in the title.
What I donāt understand is whether compressible or incompressible flows are used in the sim for calculating the dynamic pressure in order to calculate the pitching/yawing/rolling moment coefficients and resulting movement. Itās never pointed out anywhere in the documents.
compressible of course
-
What is meant by not modeling air is that there is no n-element tracking of the air medium. Computations are done with the assumption that the airās properties are irrespective to previous interactions with objects. Itās not possible to for example have an airplane change the air and that changed air to affect another airplane. There is no air medium which can take on a history or have a memory. If you want an airplane behind another airplane to feel wake turbulence then it must be done directly from plane to plane, not plane-air-plane.
well you are actually wrongā¦
in BMS, wake turbulence is done via AC - AIr - AC ā¦
how ?
i have created a AIR rotation behind each AC and the physical model is reacting to the roation of the air around it ā¦ how ? well easy , the fact that the AIR is rotating makes it variable in force and direction, which is taken into account into the model as a varation of local parameters (like AOA , speeds , sideslip angle etc etcā¦)
so YES there is AIR in BMS, if you read the BMS flight models articles you will see that the improvment of AIR modelisation has taken several monts/years
without AIR, turbulence modeling would have been impossibleā¦
i think you need to go flying back on FF or AF just to remind you what it wasā¦then you will realize in BMS you fly IN AIR compared to the others
-
Mav, I was under the impression (engendered by the articles available for download from this very website) that rather than each packet of air being independently calculated for each quanta of time, instead forces are applied to each control surface individually and the vectors are summed to determine the course of the aircraft, with forces varying according to a number of factors.
this I believe is what molni was getting at, rather than attempting to point out that no real air was harmed in the making of this sim.
I trust my impression was not in error? if so, I might recommend the updating of the articles on the flight model.
I hope this is not the case, as they were a most entertaining and informative read.
there are two aerodynamics models developped for BMS, one local (the one you described) used only by A10 in BMS and the other Global (the NASA one) used by F16 - AFM.
as said in the article, there is no NS calculation indeed , but saying there is no AIR is wrongā¦without AIR modeling , there would be ZERO force applied ā¦
for instance Local Aero modeling models AIR on each part of the ACā¦with local AIR parameter
you guys believe Navier Stokes calculation IS the answer ā¦but for a flight sim this is COMPLETELY uselessā¦why calculating NS on a surface in real time when you know ALREADY (by Navier Stoke calculation done not in real time or by reverse engineering , or wind tunnel testing) what is the RESULT you need to simulateā¦
you wont get better result by calculing in real time NS on each surface rather than calculating (or testing) the same prior simulation, translate the result into tabular data and use the data in real time
i still dont understand why simmers dont get this pointā¦
Flow equations are used to DEVELOP an aircraft , NOT TO SIMULATE IT once it already flies and been tested
-
and actually to answer the original question
YES i could improve the AIR modeling of BMS by making the airspeed (which i did already for jetwash) and pressure different behind a ACā¦which will make immediatly the AC behind react accordinglyā¦but jeeeeeez , i dont want to do that
-
-
X-Plane has no tabular (already stated in the thread). And Iāve heard of reports that it only works fast enough for prop aircraft, travelling at rather subsonic speedsā¦ That they fall back to tabular once aircraft gets too fast, since they canāt integrate with delta small enough for the results to make sense anymore.
Note: I have tons of respect for Laminar devs, especially Ben Supnik, whoās a very helpful and nice fellow, so this isnāt my way of dissing their research either.
Another question - is landing gear so resistant because you integrate with dt too high? Or is there another reason?
-
you guys believe Navier Stokes calculation IS the answer .
I know that is not the answer, because I also learned some fluid mechanics.
I tried to explain some Hungarian poeple if the physical code is good and your data is good concerning on thrust, weight, sizes and aero coefficients, even commercial flight sims are not so bad concerning the simulated Ps, but they did not believe meā¦
(I just do not know where came the date which are outside the normal operational range of an AC. I think here about the extreme big transiend AoA and stall cases.)
-
Isnāt data from TP-1538 up to alpha 90 deg enough?
-
well you are actually wrongā¦
in BMS, wake turbulence is done via AC - AIr - AC ā¦
how ?
i have created a AIR rotation behind each AC and the physical model is reacting to the roation of the air around it ā¦ how ? well easy , the fact that the AIR is rotating makes it variable in force and direction, which is taken into account into the model as a varation of local parameters (like AOA , speeds , sideslip angle etc etcā¦)
so YES there is AIR in BMS, if you read the BMS flight models articles you will see that the improvment of AIR modelisation has taken several monts/years
without AIR, turbulence modeling would have been impossibleā¦
i think you need to go flying back on FF or AF just to remind you what it wasā¦then you will realize in BMS you fly IN AIR compared to the others
Teeeeeechnically Iām not wrong because I started my sentence with āwhat is meant byā¦ā so really I was just clarifying a statement made by someone else. I never actually said any of my clarification truthfully applied to BMS.
I still think you misunderstand my point though. What happens if you instantaneously delete the aircraft creating the wake. Does the trailing aircraft feel the wake still or does it disappear with the source? To put it another way can the localized quality or air (wake, de-oxygenation, etc.) exist standalone separate from its source?
Iām not talking about the quality of the simulation. Itās entirely possible to have self-interaction through the medium well modeled without ever modeling air as a medium for the exchange. If one knows that this mach, AoA, beta, etc produces this effect then the inter alia medium of air can be assumed and bypassed. The results are very difficult if not impossible to distinguish in practice.
-
For final clarification:
Wake is not modeled in BMS , only JetWash is modeled
Wake implies that the turbulence vortex are living a long time after their generation and are impacted by wind
like this for instance:
-
Is ground effect modelled?
-
yes, heat and mechanical
-
Is ground effect modelled?
yes ground effect is modeled
Convective turbulences and mechanical turbulences as well.