Problems with the new props

[Aegolius13]

Like many of you, I'm messing around with the brand-new propellers and engines.  I have an ugly but functional 2-engine plane - pic below. 

After much trial and error, I have everything set so that the plane goes the right direction when I hit the throttle.  The engines and propellers are symmetical and both spinning the same direction; the propellers are "deployed," and their yaw/pitch/roll options are inactive.

The plane does OK on the ground, but after I take off, I get all kinds of strange yaw and roll, always in the same direction.  I have plenty of control authority to counteract it manually.  But SAS seems to have no idea what to do, and if anything makes it worse.  If I turn SAS off, I can mostly correct it by setting some yaw and roll trim, but the thing still won't go straight.   

Once I turn off the engines, the plane appears to fly correctly, both via the SAS and manually (if I set the trim back to neutral)

Really not sure why I'm getting any net roll or yaw since the two engines should be cancelling each other out, and everything else should be symmetrical

Any thoughts? Is this just one of those things where the game calculates physics for the two engines slightly different for the two engines?

 

 

 

[Guest]

Those are torque effects. Make the motors spin in opposite directions (with prop pitch mirrored to match) and they’ll go away.

You will also need to adjust pitch to match your speed. Props are more complex than jets.

Edited July 12, 2019 by Guest

[mystifeid]

Engine 1 - clockwise; Props - clockwise variant and deploy direction normal.
Engine 2 - counter-clockwise; Props - counter-clockwise variant and deploy direction inverted.

Don't forget to set the number of attachment nodes on the engines before adding the props.

I found it easier to add engines to nacelles one at a time. Adding them in pairs with symmetry and then removing them from symmetry seemed to cause some strange things to happen. For example - reverting to the hangar would sometimes see the engines back in symmetry again.

With SAS engaged, this four engine plane takes off in a gentle climb with no input required (hands free). No roll or yaw input required to maintain level flight.

Cruises at 225m/s at 3000m with prop authority at 72% and engine torque at 25%

The idea was to take off and fly directly east over the ocean until the next land mass was reached. And it made it with a reasonable amount of fuel to spare. I figured if a plane-building novice like me could do it in a reasonable time frame then the new parts were going to be ok.

Edited July 12, 2019 by mystifeid

[PhylumCnidaria]

In my experience, another weird thing that props do is put out the same level of thrust regardless of throttle. I thought that I had set up my throttle wrong (like it was either off or on 100%) but I didn't. I could reliably throttle down to like 5% and sip on fuel while getting the exact same acceleration. Also blade pitch still eludes me, it doesn't seem to matter what the pitch of the blades are for some reason. These new props act really funky compared to the old control surfaces. 

[Guest]

13 minutes ago, PhylumCnidaria said:

In my experience, another weird thing that props do is put out the same level of thrust regardless of throttle. I thought that I had set up my throttle wrong (like it was either off or on 100%) but I didn't. I could reliably throttle down to like 5% and sip on fuel while getting the exact same acceleration.

If your engine is much more powerful than it needs to be, then that's entirely as it should be. On a propeller plane, the motor's job is to keep the propeller spinning, the amount of thrust produced depends on effective blade pitch, which is a function of RPM and forward speed. This means that you need a coarser pitch at higher airspeeds and a finer pitch to get moving/accelerate/take off. If your motor has insufficient torque it won't be able to overcome the air resistance of the propeller spinning, your RPM will go down, and your thrust will drop. You will have maximum efficiency at a fairly fine pitch, which should allow you to reduce throttle. But if you can go down to 5% before your RPM drop it means your motor is way overpowered for the job.

As to blade pitch, as stated the optimal pitch depends on your airspeed. A fixed-pitch propeller will have a thrust curve: it starts out pretty low at zero airspeed, rises as you build speed, then drops again as you near the aircraft's maximum speed. A variable-pitch propeller lets you push the peak of the thrust curve forward by coarsening the blade pitch, until one of three things happen: the drag produced by your airframe equals the maximum thrust produced by your propeller, your propeller's air resistance overcomes your motor's torque, or your blade tips break the sound barrier (I don't know if KSP models this though). 

So, again: if you're making a fixed-pitch propeller, experiment to find a blade pitch that lets you get rolling and produces enough thrust at your lift-off speed to get you into a climb, and if you're making a variable-pitch propeller, taxi at a fine pitch, accelerate, take off, and climb at a slightly coarser pitch, and cruise at a much coarser pitch. Then lower your throttle as much as you can while maintaining RPM.

[Aegolius13]

2 hours ago, Brikoleur said:

Those are torque effects. Make the motors spin in opposite directions (with prop pitch mirrored to match) and they’ll go away.

You will also need to adjust pitch to match your speed. Props are more complex than jets.

 

1 hour ago, mystifeid said:

Engine 1 - clockwise; Props - clockwise variant and deploy direction normal.
Engine 2 - counter-clockwise; Props - counter-clockwise variant and deploy direction inverted.

Thanks.  I had a nagging feeling they needed to go opposite directions,  but thought every other combination of engine and prop direction failed.   Must have messed up some other detail,  though. 

[PhylumCnidaria]

2 hours ago, Brikoleur said:

If your engine is much more powerful than it needs to be, then that's entirely as it should be. On a propeller plane, the motor's job is to keep the propeller spinning, the amount of thrust produced depends on effective blade pitch, which is a function of RPM and forward speed. This means that you need a coarser pitch at higher airspeeds and a finer pitch to get moving/accelerate/take off. If your motor has insufficient torque it won't be able to overcome the air resistance of the propeller spinning, your RPM will go down, and your thrust will drop. You will have maximum efficiency at a fairly fine pitch, which should allow you to reduce throttle. But if you can go down to 5% before your RPM drop it means your motor is way overpowered for the job.

As to blade pitch, as stated the optimal pitch depends on your airspeed. A fixed-pitch propeller will have a thrust curve: it starts out pretty low at zero airspeed, rises as you build speed, then drops again as you near the aircraft's maximum speed. A variable-pitch propeller lets you push the peak of the thrust curve forward by coarsening the blade pitch, until one of three things happen: the drag produced by your airframe equals the maximum thrust produced by your propeller, your propeller's air resistance overcomes your motor's torque, or your blade tips break the sound barrier (I don't know if KSP models this though). 

So, again: if you're making a fixed-pitch propeller, experiment to find a blade pitch that lets you get rolling and produces enough thrust at your lift-off speed to get you into a climb, and if you're making a variable-pitch propeller, taxi at a fine pitch, accelerate, take off, and climb at a slightly coarser pitch, and cruise at a much coarser pitch. Then lower your throttle as much as you can while maintaining RPM.

That actually makes a lot of sense, thank you! One thing though, how exactly are variable pitch propellers made? When I used control surfaces I just deployed them and undeployed them and set the authority so that it would make the blade pitch more coarse for higher top speed. That doesn't really seem to work with these new propeller blades and I don't really know why...

[Guest]

31 minutes ago, PhylumCnidaria said:

That actually makes a lot of sense, thank you! One thing though, how exactly are variable pitch propellers made? When I used control surfaces I just deployed them and undeployed them and set the authority so that it would make the blade pitch more coarse for higher top speed. That doesn't really seem to work with these new propeller blades and I don't really know why...

I'm currently on holiday without my kerbal-puter so I haven't actually used the new parts (and won't until the end of the month), but I believe you can bind the deployment angle to an axis in the action groups, so e.g. up/down or forward/back increases/decreases it (i.e., the pitch). 

[mystifeid]

9 hours ago, PhylumCnidaria said:

That doesn't really seem to work with these new propeller blades and I don't really know why...

Have you assigned keys for the custom axis groups in the game settings? (Under Input -> Vessel -> Axis Actions). I use numberpad keys / and *, 8 and 9, 5 and 6, 2 and 3 for the four groups. Assigning the prop authority limit to custom axis group 1 lets me increase/decrease the prop authority by using the / and * keys.

Not sure how much fuel/weight anyone else is carrying but the plane above has 880 units of LF - more than enough to get me over the ocean to the east of KSC.

Previously I had just been using engine torque limit and the prop authority limit to maximize flight distance - above with the 225m/s speed, each engine was reporting fuel consumption of around 98 mU/s.

So now I'm using RPM limit as well and am flying faster while using less fuel:

Speed - 255m/s (+30m/s)
RPM - 285
Torque - 35%
Prop authority - 42%

Each engine is now reporting using around 81 mU/s. That represents a significant increase in flight distance. So it's definitely worth playing with the settings.

Edited July 13, 2019 by mystifeid

[Aegolius13]

Does anyone know what the "invert" option on the rotors/engines does, and how it's different than changing the rotation direction?  I think my original problem trying to mirror might have had something to do with that.

[Geschosskopf]

17 hours ago, Aegolius13 said:

Does anyone know what the "invert" option on the rotors/engines does, and how it's different than changing the rotation direction?  I think my original problem trying to mirror might have had something to do with that.

In the action group editor, checking "invert" changes the direction the axis moves.  So for instance, if you invert main throttle, the plane spawns with the throttle set to 0.  However, this will be interpreted as full throttle so the plane will spawn with full throttle.  Hitting LSHF, which normally increases throttle, will instead decrease it.

To make counter-rotating, separate engines (with 1 engine on each wing) come out right, do the following:

1. Attach the 1st engine to the engine mount of 1 wing with radial (VAB-style) symmetry set to 1 instance so you only get the 1 engine.  You need to use radial symmetry here for step 2 to work.  It's best to start with the engine on the left wing.
2. Still in radial (VAB) symmetry, attach however many blades you want to the 1st engine's hub.
3. Switch to mirror (SPH-style) symmetry, change instances to 2.
4. Grab the 1st complete engine assembly (rotor plus blades) and put it back.  This creates a mirrored duplicate on the other wing engine mount.  The advantage of doing things this way is that it mirrors the direction of the blades on the 2nd engine, so you don't have to go back and change them later to spin the opposite direction.
5. Right-click on the 2nd engine (right wing) and select "Remove from symmetry", then reverse direction.  Do NOT remove its blades from symmetry.  

So now, when viewed from the front, the engine on the left wing will spin clockwise and the engine on the right wing will spin counter-clockwise.

To make a 2-engine push-pull plane (engines fore and aft), it's a bit different.

1. Build the front engine as in steps 1 and 2 above.
2. ALT-click on the complete front engine assembly to clone it, rotate it 180^, and attach the clone on the rear end.  Rotating the engine around backwards takes the blade direction with them, so again you don't have to switch the blades around.  Also, as the 2 engines where not installed with symmetry, you don't have to remove the rear engine from symmetry, or even change its direction of rotation.
3. What you DO have to change, however, is the pitch of the blades.  They have to be angled the opposite direction as the front blades.  This includes any deployment authority you have set up for variable pitch.

NOTE:  There is some evidence that KSP models air being accelerated back past the plane by forward props.  Thus, the rear prop could well need slightly more pitch than the front prop.

[Aegolius13]

On 7/13/2019 at 2:13 PM, Geschosskopf said:

In the action group editor, checking "invert" changes the direction the axis moves.  So for instance, if you invert main throttle, the plane spawns with the throttle set to 0.  However, this will be interpreted as full throttle so the plane will spawn with full throttle.  Hitting LSHF, which normally increases throttle, will instead decrease it.

I think we're talking about different instances of "invert."  The one I'm referring to is "Invert Direction" on the right-click menu for the part in build mode - see picture below.  I did some more testing, and as far as I can tell, it does reverse the direction the motor rotates - so "clockwise + normal" and "counterclockwise + invert" result in the same direction.  Based on the italic text under the button, it appears that this lets you change the direction of only one engine in a symmetrical set -- allowing you to do counter-rotating props without removing the engines from symmetry for other purposes.   In that case, seems pretty handy!  [EDIT - on second thought, this might be problematic for reversing the propeller direction.  But could be useful for other contraptions - a pitching machine comes to mind.]

Guessing my original error involved using the two commands inconsistently, maybe in combination with the props ending up extending in the right direction.  But at any rate, the plane is now working as expected with no net torque.

Edited July 16, 2019 by Aegolius13