Game 'physics' realism VS 'aircraft systems' realism

[OzWhitey 3,353]

12 minutes ago, Stearmandriver said:

Unfortunately not, because all our performance handbook tables and calculator assume the ubiquitous engine failure at V1.  I'm pretty sure there's no data I have access to that shows unfactored data with all engines operating. I'll take a dig through later though... maybe there is a table in there. We would just never use it. 

Well, I’d certainly be interested in any insights you have.

But I’m a scientist, so I was thinking more along the lines of generating some new experimental data. All I need is a full 737, a short runway, and someone willing to give a flaps 0 takeoff a really good go. Hey, it’ll be fun. And also…science.

😉

Edited May 13, 2022 by OzWhitey

[sd_flyer 6,040]

5 hours ago, Doug47 said:

That’s not correct at all. Their take off data, from the testing, their programs they use for flap settings, V speeds etc would tell them if it could. 
 

A quick flip through google (not a super valid source I know)shows for Flap 5, standard day a 737-700 would be limited to around 125,000lb (57,000kg) using a 5000ft runway. 
 

Another source says at 154,000 lbs it needs 5500ft. 
And I’m guessing that’s at flap 5 as well. 

So heavier, at zero flap? It wouldn’t easily take off at all. You’d be hard pressed to believe 5000ft would be enough 

 

You are referring to a data that come out of certification with huge safety margins. The test process however can push airframe way beyond limits. 

[sd_flyer 6,040]

5 hours ago, Jetman67 said:

Great pilot and good enough to not get near an assymetric spin or would have been lights out most likely

Bob Hoover not just a good pilot he is legendary pilot among Chuck Yagers and other!

 

[sd_flyer 6,040]

 

B-707 barrel roll !

 

[sd_flyer 6,040]

53 minutes ago, OzWhitey said:

Well, I’d certainly be interested in any insights you have.

But I’m a scientist, so I was thinking more along the lines of generating some new experimental data. All I need is a full 737, a short runway, and someone willing to give a flaps 0 takeoff a really good go. Hey, it’ll be fun. And also…science.

😉

That someone you need is called a test pilot. LOL My dad was one ! 😎

[Stearmandriver 2,279]

1 hour ago, OzWhitey said:

Well, I’d certainly be interested in any insights you have.

But I’m a scientist, so I was thinking more along the lines of generating some new experimental data. All I need is a full 737, a short runway, and someone willing to give a flaps 0 takeoff a really good go. Hey, it’ll be fun. And also…science.

😉

Yeah I mean that's certainly been done before by a test pilot (on a conservatively long runway of course), but this is the type of data that's not typically for public consumption. 

The scientists I know would typically like to have some data to base a hypothesis on before conducting an experiment, but I'm sure your way works too.   😁

[Bill Griffith 160]

5 hours ago, ryanbatc said:

How would we even know?  Most of us are using basic or semi-basic flight controls, sitting in front of a monitor.  I quit chasing trying to get everything right from a feeling standpoint a long time ago...  There are too many user variables.

    Years back a friend of mine who's father flew P-51's in Europe had him try a P-51 in a flight sim.  He then asked his father if that felt like a P-51.  His father's reply was, "What do you mean, feels like a P-51?  It's a box on the desk!"   That is where we are.  😊

[mrueedi 702]

5 hours ago, Stearmandriver said:

Unfortunately not, because all our performance handbook tables and calculator assume the ubiquitous engine failure at V1.  I'm pretty sure there's no data I have access to that shows unfactored data with all engines operating.

Here is the data, nicely and logically dissected into the relevant pieces. It confirms your conclusion:

19 hours ago, martinboehme said:

An interesting discussion. Some have stated that it’s impossible to answer the question whether this takeoff would be possible in reality, but I believe that at least Boeing should be able to answer this question based on their flight test data, and I think we can even get some of the way using only publicly available information.

There are two questions we need to answer:

1. What speed will the aircraft reach by the end of the runway?

2. At that speed, how much lift will the aircraft produce in the maximum attitude that can be achieved on the ground (i.e. tail strike attitude), in a clean configuration?

If the lift is greater than the weight of the aircraft, the takeoff should be possible (assuming no obstacles, perfect execution etc.), though of course not necessarily legal.

Boeing’s test data should be able to answer these questions.

Regarding question 1, they will have performed extensive takeoff acceleration tests. While they will likely not have performed these beyond MTOW, the data for these tests should extrapolate reliably beyond MTOW: Aerodynamic drag is not affected by the TOW, and there’s a simple inverse relationship between mass and acceleration. Tyre rolling friction will increase with TOW, but it’s a simple physical principle that should extrapolate reliably, and its influence is likely small anyhow.

Regarding question 2, the answer to that does not depend on TOW, and Boeing will certainly have precise lift polars for every configuration.

We’re not Boeing, but let’s see how far we can get with publicly available data.

First, let’s set some parameters. I’m going to assume a sea level ISA day with no wind. I don’t know the TOW that the OP used, so I’m going to assume MTOW.

From http://www.b737.org.uk/techspecsdetailed.htm, we find the following relevant data for the 737-700:

Static thrust 89.0 kN per engine, so 188.0 kN total
MTOW 77564 kg
Wing area 124.58 m^2

(Static thrust data is often quoted for sea level ISA conditions, so I’m assuming that’s the case here.)

The runway length at EGLC is 1508 m.

Let’s answer question 1 first. In doing so, we’ll have to make some simplifying approximations:

• We will neglect drag and tyre friction
• We will neglect the increase in thrust due to the ram effect, i.e. we will assume that thrust remains constant at the static thrust

These two approximations have opposite effects, but all in all, I expect that we will be overestimating the speed. Acceleration tends to reduce during the takeoff roll, which implies that drag and tyre friction have a larger effect than the ram effect. What this means is that the calculation will be too optimistic: It may predict that a takeoff is possible when in reality it is not. Conversely however, if the calculation says the takeoff is not possible, we can be pretty sure that it would not be possible in reality.

With that, let’s plug in some numbers and see what we get.

Acceleration is

a = F / m = 178 kN / 77564 kg = 2.3 m/s

This gives us the following speed at the end of the runway:

v = a sqrt(2s / a) = 2.3 m/s * sqrt(2 * 1508 m / 2.3 m/s) = 83.3 m/s

That’s 162 knots, which seems like it might be barely enough.

Let’s continue with question 2. I’m going to assume that the 737-700 can reach clean CLmax in the tailstrike attitude. Unfortunately, it’s hard to find a clean CLmax for the 737-700, so I’m going to use the typical CLmax ranges found in this source; again, to be optimistic, I’ll use the top end of the range, giving me a CLmax of 1.8.

With an air density in sea level ISA conditions of rho = 1.225 kg/m^3, the lift we can achieve becomes

L = CL rho v^2 / 2 A = 1.48 * 1.225 kg/m^3 * (83.3 m/s)^2 / 2 * 124.58 m^2 = 783.6 kN

The weight of the airplane at MTOW is

Fg = m g = 9.81 m/s^2 * 77654 kg = 761.8 kN

So we’re achieving just slightly more lift than the aircraft’s weight, but keep in mind that we’ve made various optimistic assumptions and approximations. Without these, it’s possible we wouldn’t quite be able to achieve the required lift.

We could, however, make the conditions more favorable than sea level ISA, nil wind. Lift increases with the square of the airspeed, so a ten-knot headwind, say, is going to help a lot. We could also lower the temperature, which would give us both more thrust and more lift. Changing both of these environmental conditions could be enough to compensate for the optimistic assumptions we’ve made. (@Bad_T, I'd be interested to know what the weather conditions were for your test.)

In the end, the results are close enough that we can't really make a definitive statement either way, but I'd say we can conclude that getting a 737-700 out of EGLC at MTOW is at least not wildly unrealistic.

Some remarks from me about this calculation:

19 hours ago, martinboehme said:

Unfortunately, it’s hard to find a clean CLmax for the 737-700, so I’m going to use the typical CLmax ranges found in this source; again, to be optimistic, I’ll use the top end of the range, giving me a CLmax of 1.8.

With an air density in sea level ISA conditions of rho = 1.225 kg/m^3, the lift we can achieve becomes

L = CL rho v^2 / 2 A = 1.48 * 1.225 kg/m^3 * (83.3 m/s)^2 / 2 * 124.58 m^2 = 783.6 kN

You say CL = 1.8 but use 1.48? Why not use this source:

Solved For the airfoil polar diagram of a Boeing 737 shown | Chegg.com

It suggests that 1.6 should be fine. It is also flaps up. It would give L = 847 kN, which is a solid reserve over 761 kN.

The only thing I am not sure, is whether drag in the second half of the take of run would start having a considerable impact.

Edited May 13, 2022 by mrueedi
typo

[Fielder 3,663]

The on the ground handling of tail draggers in MSFS is not very realistic at all. On one of the Q&A's Seb said the contact area between tires and pavement in MSFS is based on a point, not based on an area. And so the rolling friction is sometimes wrong and not based on realistic physics, and is sort of hacked. That is what Seb said.

They are working on a more advanced model which will produce more accurate results, but it is complicated and not easy, he said.

I don't see any other real odd behavior (except all planes tend to float a little bit too much above the runway not enough tendency to drop). But only a very little off. Tail draggers dancing the twist down the runway is more than just a little bit wrong.

 

[gb09f 298]

On 5/11/2022 at 1:33 PM, abrams_tank said:

In fact, flightdeck2sim, another real life 737 pilot, said that the PMDG 737 for MSFS handles and performs better than the PMDG 737 for P3D on his Youtube live stream yesterday.

He did say that, but he rates the freeware Zibo 737 as having the most realistic flight model. 

[Stearmandriver 2,279]

19 minutes ago, Fielder said:

The on the ground handling of tail draggers in MSFS is not very realistic at all. On one of the Q&A's Seb said the contact area between tires and pavement in MSFS is based on a point, not based on an area. And so the rolling friction is sometimes wrong and not based on realistic physics, and is sort of hacked. That is what Seb said.

They are working on a more advanced model which will produce more accurate results, but it is complicated and not easy, he said.

I don't see any other real odd behavior (except all planes tend to float a little bit too much above the runway not enough tendency to drop). But only a very little off. Tail draggers dancing the twist down the runway is more than just a little bit wrong.

 

This is all true; however as the PMDG 737 is using a custom ground physics model as well as flight model, it probably doesn't apply to the question in this thread. 

[abrams_tank 7,397]

11 minutes ago, gb09f said:

He did say that, but he rates the freeware Zibo 737 as having the most realistic flight model. 

He also said the PMDG 737 has better systems than the Zibo.

It seems like a wash for me. 

[flying_carpet 657]

I'm wondering why the title of the thread "GAME 'physics' ..." didn't cause an outcry 😃.

 

1 hour ago, mrueedi said:

You say CL = 1.8 but use 1.48? Why not use this source:

Respect, someone had a closer look 😉. That's what I also found after checking the calculation. At first, it looks very impressive 😉, but only by checking yourself you can find out wether it's correct.

Apart from that and additionally ...

19 hours ago, martinboehme said:

Static thrust 89.0 kN per engine, so 188.0 kN total

Acceleration is

a = F / m = 178 kN / 77564 kg = 2.3 m/s

  

1 hour ago, mrueedi said:

The only thing I am not sure, is whether drag in the second half of the take of run would start having a considerable impact.

I'm "afraid" it does: drag rises quadratic, i.e. at 162 kn it's 4 times higher than 81 kn.

 

39 minutes ago, gb09f said:

He did say that, but he rates the freeware Zibo 737 as having the most realistic flight model. 

Lese majesty 😄

Edited May 13, 2022 by flying_carpet

[OzWhitey 3,353]

7 hours ago, Stearmandriver said:

Yeah I mean that's certainly been done before by a test pilot (on a conservatively long runway of course), but this is the type of data that's not typically for public consumption. 

The scientists I know would typically like to have some data to base a hypothesis on before conducting an experiment, but I'm sure your way works too.   😁

Mate, I don’t tell you how to fly the plane, so perhaps don’t tell me how to do the science thing, OK?

But seeing you think “your scientists” are all superior:

Of course we gathered data as the first step. Read OP’s post, it’s all published there. We’ve already tested the hypothesis in one of the best simulators available. The tests are consistent - you’ll be fine with the EGLC flaps zero takeoff,probably. 

We could keep faffing around collecting data in the sim forever - but the time has come to move on to human trials, out in the wild. So load that plane up to the gunwales, find an excuse to land at London City and then - prepare for takeoff!

😉

 

[Glenn Fitzpatrick 3,798]

OK - you have 4900 feet at EGLC.

These idiots from Southwest landed at the wrong runway (3700 feet) and managed to get out again.