Stock Turboprop Endurance Record

[Pds314]

3 hours ago, Gman_builder said:

I just threw together a engine in 30 minutes to an experiment and it ended up doing 50 rad/s at half throttle! WOW! I have got some work ahead of me tonight!

EDIT: Makes 40 rad/s before the axle comes out of the frame for whatever reason. But with a small prop on the front it does 7 rad/s  how does that even work

Basically, the problem is likely that you were running it unloaded. Running a turboshaft unloaded is a very bad idea (or very good if you want explosions). The thing is that shaft drag is extremely tiny when it is operating correctly.

So power is conserved, and since:

Power input = angular velocity * torque.

Power output = power lost to prop + power stored as angular velocity + power lost to shaft + power stored as vibration.

Power input = power output.

So running it without a prop means that it will continue to accelerate more and more until the shaft is absorbing the entire output of the engine, or, more likely for engines of significant power, something explodes, collides with something it shouldn't, or vibrates apart.

Practically speaking and for engines of good bearing design, there is no closed-form solution for the speed a running engine with no prop will reach. Power in keeps increasing linearly with angular velocity, while vibrations are subject to instability only at high speeds. This means that for a running engine to avoid runaway rotation, the power dissipation must increase at a rate faster than linear, and what that means is that most engines will continue to accelerate until they explode.

3 minutes ago, Azimech said:

Inferiour? Maybe. But it's got to do with RPM, prop diameter and number of blades.

That too. I'm having to give it pretty high pitch to reach those speeds, which means the efficiency of the blades probably isn't phenomenal either. The RPM and prop diameter are really the issue since they drive the advance ratio up well beyond what's efficient, I think. Keep in mind, however, that prop diameter varies considerably in flight due to rotation rate.

Edited July 12, 2016 by Pds314

[Azimech]

Guys, let's copy and paste our conversation in the main turboshaft topic so other people can be helped with the theory.

 

[Pds314]

2 hours ago, Azimech said:

Guys, let's copy and paste our conversation in the main turboshaft topic so other people can be helped with the theory.

 

Agreed. Which one is currently active?

[DoctorDavinci]

Perhaps you should add a note in the OP about changing the physics config as this is the first I've heard about the rpm being limited

[Azimech]

3 hours ago, Pds314 said:

Agreed. Which one is currently active?

 

[Gman_builder]

12 hours ago, Pds314 said:

Interesting. One thing to point out is that encountering the sound barrier might reduce prop efficiency. The effects aren't fully simulated in KSP, but the efficiency of a supersonic propeller is likely considerably lower than that of a subsonic one. Say you have a prop turning at 40 rad/s as it approaches the sound barrier, and the blades are 4 meters from the hub, you have a horizontal speed component of 160 m/s, and let's say a linear speed component of 310 m/s.

Your prop blades are now moving 349 m/s. This means they are supersonic. This considerably reduces their L/D, yet also requires the advance ratio to be more than 2.0 to generate meaningful thrust. So let's say your prop is getting an L/D of 4? That means that in order for the prop to overcome its own drag, it must have an advance ratio of less than 4, but in order to generate that L/D of 4, it needs at least several degrees of AOA on the airstream. There's a very slim margin where the prop is generating useful thrust that's also capable if overcoming the prop's own drag. Add in the drag of the aircraft and you start to see why supersonic props are mindbogglingly inefficient.

 

Then again, you could fix some of these things. Change the required advance ratio by spinning the prop faster (dangerous) or making it wider (draggy). Improve supersonic aerodynamics by using swept props (inefficient at low speeds, may be hard to control pitch, may cause structural problems). Perhaps a 50 rad/s prop with swept blades centered 6 meters out would have more promising supersonic behavior.

Ya in my run at 224 m/s the prop was going well over supersonic at 44 rad/s but it kept accelerating until either the engine broke or the game crashed. But my point is that at higher altitudes your engine could be producing more thrust at higher efficiency by using panthers. An engine I am testing now has a single row of 8 panthers in a 4.5m diameter frame with a 3 bladed static prop similar to the one off the F8F or DC-7 and a 3.5 fairing as the nose cone. the engine can easily do 35 rad/s or higher but when it is on a plane I seem to get a strange uncontrollable upward pitch motion when the engine is operating somewhere above 30 rad/s I havn't pinpointed the exact RPM. But the engine itself Is extremely stable and reliable even through hard maneuvers besides that strange pitch up thing. If you have any ideas on that i'd love to hear them. That plane does have considerable better efficiency when flying at higher altitudes and speeds in excess of 100 m/s.

11 hours ago, Azimech said:

Guys, let's copy and paste our conversation in the main turboshaft topic so other people can be helped with the theory.

 

Lol that topic has been dead for months. I havn't seen any activity on it besides my own in forever.

[Azimech]

12 minutes ago, Gman_builder said:

Lol that topic has been dead for months. I havn't seen any activity on it besides my own in forever.

Check again ;-)

[Gman_builder]

1 minute ago, Azimech said:

Check again ;-)

ya I just did. I stand corrected.

[Gman_builder]

6 hours ago, DoctorDavinci said:

Perhaps you should add a note in the OP about changing the physics config as this is the first I've heard about the rpm being limited

changed

[DoctorDavinci]

I've noticed that when I increase the AoA of my propellers that my rad/s go down

 

[Gman_builder]

6 hours ago, DoctorDavinci said:

I've noticed that when I increase the AoA of my propellers that my rad/s go down

 

Ya that's kind of been known for, like, forever with static engines. If your flying, the effect of increasing prop pitch depends on how fast your going. Ideally, at low speeds you want lower pitch, and high speed you want highest pitch to maintain RPM.

I.E.  

Low speed = low enough pitch to allow high RPM, but at the same time high enough not to overspeed the engine

High speed = High enough pitch to allow high RPM, but at the same time low enough to not stall your prop.

 

[Cunjo Carl]

Oh! If it's propeller blade angles we're interested in, I can add something to the table. I've calculated quantitatively the L/D for the wings. It turns out all the wings behave the same L/D-wise (much to my surprise), and are also the same lift/mass-wise. The only exception is the basic fins with 20% better lift/mass somehow. That said, longer propeller wings will still catch more air, because their effective velocity is still determined by its size (2pi*r*spin-speed). Besides that, a wing's a wing, which is nice to know.

Also, just to make sure we're on the same page, the "lift" below correlates directly to the force of the turboprop, and the "drag" is one of the turboprop's key sources of friction.

These values were plotted for a static (0m/s) test, but if people were interested, I could drum this up in a real program to include different velocities. I could also show how to turn these values directly into forces if that'd be handy.

Blue dots (upper line) is LIFT. Red dots (lower line) is DRAG. This takes into account that some of your lift is pointed counter-turn-wise and is actually acting as drag instead.

We can also take the ratio of these to obtain the commonly used L/D ratio for determining whether something is a good idea, or it'll get you stalled / make your turboprop spin too slow.

How charting KSP phsics works (my simple, homebrewed method).

Spoiler

1. Find the data

The data is written in config files using the following format

  key₁   = X₁  Y₁  I₁ O₁
  key₂   = X₂  Y₂  I₂ O₂
  key₃   = X₃  Y₃  I₃ O₃
  key₄   = X₄  Y₄  I₄ O₄

Where each key defines a point in terms of its x value, y value, incoming tangent (I), and outgoing tangent (O). A good example of these values might be mach values for x (the thing you know), drag coefficients for y (the thing you want to know), and the I and O tangents for how quickly drag coefficients change approaching and leaving this point.


2. Find some coefficients

We'll need some coefficients in a moment, so we plug in some numbers from key1 and some from key2.

  A =  2Y₁ +  O₁  - 2Y₂ + I₂
  B = -3Y₁ - 2O₁ + 3Y₂  - I₂
  C =    O₁
  D =    Y₁
 

3. Form the equation

We'll write an equation for the line that goes between the points of key1 and key2. We could try to make this equation in terms of x, but it would be really hard! So instead, first we'll make a convenience variable called u. The fancy term for it is a unitless transform, and it makes life soooooo much easier. Then we'll write an equation for y (the thing we want!) in terms of this u.

u = (x - X₁)/( X₂  - X₁ )           Eq. 1

y = Au³ +Bu² + Cu + D           Eq. 2

Cool, so if we make up an x value, and we want to know the corresponding y value we simply plug x into Eq 1, and then plug the resulting u into Eq 2 and we get our value! All's we need are those big letters A-D.
 

It's a little time consuming but pretty easy in practice. And if not, sorry I mathed all over the forum .

For actual application, it all boils down to trying things and seeing what works in the end anyways, but it can be nice to have charts like these to show you what to expect. I hope they help!

Edited July 27, 2016 by Cunjo Carl

[Azimech]

Nice one @Cunjo Carl! Would you like to share your post in this topic? It's the main topic and where most of the reading & posting is done. Lots of science guys over there :-)

 

[Gman_builder]

12 hours ago, Azimech said:

Nice one @Cunjo Carl! Would you like to share your post in this topic? It's the main topic and where most of the reading & posting is done. Lots of science guys over there :-)

 

That was you 3000th post. congrats!

[Azimech]

Just now, Gman_builder said:

That was you 3000th post. congrats!

Wow ... wasn't aware of it. Thanks!

[Azimech]

Let's revive this before it starts to smell :-)

[Gman_builder]

On ‎7‎/‎30‎/‎2016 at 1:31 AM, Azimech said:

Let's revive this before it starts to smell :-)

I agree. Working on a entry now.

On ‎7‎/‎12‎/‎2016 at 6:31 AM, Pds314 said:

Agreed. Which one is currently active?

BABY COME BACK to this topic.

DONT LET IT DIE

[Gman_builder]

I actually need a nap right now. Long day. Will work on it later though.