Mechjeb Ascent Settings with FAR?

[Agarax]

Title says it all. What are the best Ascent Guidance settings to use with Mechjeb+FAR?

Thanks!

[Hodo]

depends on the craft.

But you want to get the TWR around 1.2-1.4. And the best ascent guidance path is 20-25% with the target altitude around the goal orbit altitude.

[Taki117]

depends on the craft.

But you want to get the TWR around 1.2-1.4. And the best ascent guidance path is 20-25% with the target altitude around the goal orbit altitude.

I have had more success with 30-50%, but that might jsut be me. I do however Disable auto-turn start, and instead set it to .2km. It usually works well enough.

[Unto the Cruel]

Start your turn early (like ^ said), and make your ascent profile as gentle as possible. Outside of that, you have to find numbers that work for the craft in question, though fairings could probably help smooth out with some standardization.

[RoboRay]

It's actually more efficient to level off in the upper atmosphere (like 50km+) and burn horizontal there to continue slowly raising your Ap while also rapidly bringing up your Pe (taking more advantage of the Oberth effect) than leveling off at the Ap and raising the Pe to match it.

I usually set the guidance for a 66% or so curve that starts at 1km and ends at 50km. I shut down the engines with Ap at 100km and Pe at 20-50km, so only a tiny circularization burn (less than 100 m/sec) is needed.

[Hodo]

It's actually more efficient to level off in the upper atmosphere (like 50km+) and burn horizontal there to continue slowly raising your Ap while also rapidly bringing up your Pe (taking more advantage of the Oberth effect) than leveling off at the Ap and raising the Pe to match it.

I usually set the guidance for a 66% or so curve that starts at 1km and ends at 50km. I shut down the engines with Ap at 100km and Pe at 20-50km, so only a tiny circularization burn (less than 100 m/sec) is needed.

This is not true. It is less efficient to level off in the upper atmosphere and burn. That is like going straight up and turning 90deg then burning to the horizon. Real rockets don't do this because it is wasteful and stupid.

[RoboRay]

This is not true. It is less efficient to level off in the upper atmosphere and burn.

In KSP, the Oberth gains far exceed the miniscule drag losses above 50km.

That is like going straight up and turning 90deg then burning to the horizon.

It's pretty much the opposite of that. Ending your turn at the orbital altitude and burning horizontal there (as you suggested above) is much closer to what you describe here than what I describe.

Real rockets don't do this because it is wasteful and stupid.

It is actually pretty much how real launches go... when reaching an altitude where drag decreases to a negligible value, burn purely for horizontal velocity. Then, rise to Ap for a short circularization burn. See the Space Shuttle launch profile, for instance. It's somewhat in-between what what we are both proposing, though... with MECO and ET Sep after completing a long, fairly horizontal burn (actually slightly descending toward the end) at about 80 miles up, then coasting upwards to the desired orbital altitude for circularizing with the OMS.

Due to the tiny sizes of planets in KSP and extreme density of matter, but relatively tall atmospheres, some things work a little differently in KSP than they do in the real world. If you want to think that's "stupid", talk to Squad or stick with RSS. Proportionally (ratio of spacecraft altitude to planetary radius), what real rockets do would put them even lower in the atmosphere than 50km. But, KSP's drag forces us higher into less efficient profiles. Going higher than you need to go to escape drag is simply counterproductive, though.

Edited June 24, 2014 by RoboRay

[Agarax]

RoboRay, I'm using FAR. The ascent profile you are talking about is for stock.

Thanks to everyone for their suggestions!

[Dkmdlb]

RoboRay, I'm using FAR. The ascent profile you are talking about is for stock.

His point is its better to get horizontal at about 50km rather than about 70km+ like Taki117 said to do. This is true for both FAR and stock.

[Hodo]

In KSP, the Oberth gains far exceed the miniscule drag losses above 50km.

It's pretty much the opposite of that. Ending your turn at the orbital altitude and burning horizontal there (as you suggested above) is much closer to what you describe here than what I describe.

It is actually pretty much how real launches go... when reaching an altitude where drag decreases to a negligible value, burn purely for horizontal velocity. Then, rise to Ap for a short circularization burn. See the Space Shuttle launch profile, for instance. It's somewhat in-between what what we are both proposing, though... with MECO and ET Sep after completing a long, fairly horizontal burn (actually slightly descending toward the end) at about 80 miles up, then coasting upwards to the desired orbital altitude for circularizing with the OMS.

Due to the tiny sizes of planets in KSP and extreme density of matter, but relatively tall atmospheres, some things work a little differently in KSP than they do in the real world. If you want to think that's "stupid", talk to Squad or stick with RSS. Proportionally (ratio of spacecraft altitude to planetary radius), what real rockets do would put them even lower in the atmosphere than 50km. But, KSP's drag forces us higher into less efficient profiles. Going higher than you need to go to escape drag is simply counterproductive, though.

Real rockets do not go straight up.

I have seen one launch in real life. And if you watch this video, you can see this rocket sure as heck dont go straight up.

This Delta IV launch definately is starting its turn LONG before it hits 70km.

[Roastduck]

This is a time lapse photo of a real launch:

However, in KSP, I've found it to be most consistent and efficient (in terms of getting to my typical parking orbit of 100km) of beginning my turn at 7km. By 50-60km, depending on the lifter, I am completely horizontal, then I begin burning towards just below the horizon to continue accelerating while also keeping me apoapsis stable

[Unto the Cruel]

I've been toying with it a little bit. I've got it set to start the turn around 5 km and end it at 65km. I think the turn rate is set to about 60%, and it's working with standard rockets and the KSO stubby shuttle as well.

[Eric S]

Real rockets do not go straight up.

He understands that. He's not talking about starting the gravity turn at 50km, he's talking about finishing it there. Personally, I tend to aim at a 0 degree pitch even lower than that.

Just a breakdown of communication.

[Geschosskopf]

My own way of using MJ Ascent Guidance with FAR is as follows:

First, rocket construction.

A. Long, tall, and skinny, so the CoM is as far forward as possible. Minimize radial boosters because these move the CoM back, but they can be used if necessary. The current version of FAR isn't so picky here; not long ago, about the only way to use radial boosters was to attach them to the upper stages.

B. Big tailfeathers on the bottom stage. I usually use stock delta wings with the large control surfaces on them. FAR likes 4 of them. It's gotten better recently with using 3 fins but 4 still is preferred. You have the choice of putting them on the cardinal angles or the diagonals. If you put them on the cardinals, be sure to set 2 for pitch, 2 for yaw, and all for roll. If you put them on the diagonals, don't bother with their axes. NOTE: If you drop this stage while still in relatively thick air, you probably will need small tailfeathers (AV-8R) on the 2nd stage as well. Be sure to balance the CoM and CoL of just that part of the rocket too, just like you do for the whole thing.

C. RCS. Put an inline RCS tank just below the payload and 3 sets of quad thruster blocks at the bottom end of each stage in the stack. Activate RCS as soon as you launch. This really helps MJ keep the thing stable. It's also required to orient a long, tall rocket for the circularization burn at Ap, and if you're launching really heavy payloads. These RCS thrusters aren't for docking, they're for steering. The payload will have its own for docking if necessary.

The MJ settings:

1. Start the turn at 1-3km. This gives you enough airspeed for the control surfaces to have decent bite.

2. Use a 65% curve or so, so the beginning of the curve is very gentle and it steepens as you get higher.

3. Finish the turn at 100km so you never quite get totally horizontal in the atmosphere. If you set the turn finish at 70km, often MJ will have your rocket slightly nose-down in the upper air, which can cause problems.

4. Limit to terminal velocity (of course)

5. Avoid overheats (usually not a problem because....)

6. ...Limit acceleration to 15-20m/s. This is mostly if your 1st stage (at least, maybe 2nd as well) has a TWR > 1.5 or so. But anyway, limiting acceleration will keep the throttle pretty low once you get going, so overheats aren't ever a problem.

[RoboRay]

RoboRay, I'm using FAR. The ascent profile you are talking about is for stock.

No, it's for FAR. You would have very bad results starting your turn at 1km altitude in stock KSP.

Real rockets do not go straight up.

I have seen one launch in real life. And if you watch this video, you can see this rocket sure as heck dont go straight up.

This Delta IV launch definately is starting its turn LONG before it hits 70km.

I never once suggested going straight up. I don't know where you're getting that, but you are not reading it in my posts.

I specifically said "I usually set the guidance for a 66% or so curve that starts at 1km and ends at 50km."

How, exactly, does a turn start at 70km if it began at 1km and ended at 50km?

Edited June 26, 2014 by RoboRay

[BudgetHedgehog]

4. Limit to terminal velocity (of course)

Just want to pick up on that - that does absolutely nothing as FAR changes terminal velocity and MJ is calibrated to read the stock drag amounts, not FARs. There is a patch by sarbian that has basic FAR support, but I'm not sure how useful it actually is. Either way, you can leave Limit To Terminal Velocity on or off, it doesn't make a difference as MJ thinks that because your craft has no drag, it has an infinitely high terminal velocity.

[Hodo]

No, it's for FAR. You would have very bad results starting your turn at 1km altitude in stock KSP.

I never once suggested going straight up. I don't know where you're getting that, but you are not reading it in my posts.

I specifically said "I usually set the guidance for a 66% or so curve that starts at 1km and ends at 50km."

How, exactly, does a turn start at 70km if it began at 1km and ended at 50km?

Sorry missed that, so we were on the same sheet of music just talking about it in different ways. My fault.

[Geschosskopf]

Just want to pick up on that - that does absolutely nothing as FAR changes terminal velocity and MJ is calibrated to read the stock drag amounts, not FARs. There is a patch by sarbian that has basic FAR support, but I'm not sure how useful it actually is. Either way, you can leave Limit To Terminal Velocity on or off, it doesn't make a difference as MJ thinks that because your craft has no drag, it has an infinitely high terminal velocity.

I have to disagree with this. I think MJ's recently got a bit smarter on this than before and takes FAR into account. I seem to recall actually reading that in the MJ thread. And my observations back this up. If you have FAR, limiting to terminal velocity results in higher speeds at lower altitudes than without FAR. So the setting definitely makes a difference, even with FAR.

[RoboRay]

Sorry missed that, so we were on the same sheet of music just talking about it in different ways. My fault.

No problem, it happens... Done it myself.

Yeah, we're flying a very similar profile, except that I'm doing a somewhat tighter turn down low to level off at a lower altitude, still in the atmosphere but high enough that the residual drag is insignficant. This expends less energy purely to gain altitude, incurring reduced gravity losses, and extracts a little more energy from the Oberth effect by conducting the horizontal acceleration slightly deeper in the gravity well.