Thrust control to follow various parameters

[Starlionblue]

Good morning.

I'm seemingly obsessed with Eve ascents. Constantly trying to find the lightest weight vehicle.

I don't hand fly. Mostly I use MechJeb. However one of the problems I face is suboptimal thrust control. MechJeb allows setting of a dynamic pressure (Q) limitation on thrust, which is good for the initial ascent. However later I have a feeling that higher thrust is better due to an increase in terminal velocity.

I would like to be able to control thrust by matching vessel velocity (EAS) to estimated terminal velocity, both numbers found in Aero GUI from the debug menu. Another option would be thrust to be limited by a specific local Thrust to weight ratio. Say TWR reaches 2.0. At that point thrust stays at 2.0 and does not go higher.

Is there any add-on that allows setting thrust to follow various parameters like this? Alternatively, perhaps someone would like to take the challenge on?

Kind regards.

Edited June 30, 2017 by Starlionblue

[artwhaley]

Mechjeb used to have a 'limit to terminal velocity' setting....  but it was considered deprecated I believe?  You could perhaps compile an old version of mechjeb against the current KSP version?   Or....  is this something the mechjeb scripting module could be used for?   Quite possibly?   

What you want to do would be pretty trivial to accomplish in Python via KRPC...  but...  if you don't know Python...getting everything set up and learned to the point that you're READY to write that trivial program is...  non-trivial.  

[Starlionblue]

Yeah I don't know Python.

Setting a Thermal limit would also work I think. Like "don't go over x% of the max temperature for the most critical part."

[artwhaley]

I absolutely am not volunteering @linuxgurugamer , one of the hardest working modder I know, for more work...  but...  if you take a look at this thread -  

If they DON'T have a part that responds to temperature, I suspect that it's something that might get added to the eventual todo list there if you explain exactly what you'd want it to do, in context of the other parts already in that mod?

That said...  and I'm not a fluid dynamics expert...   Isn't terminal velocity essentially a property OF dynamic pressure?   If the dynamic pressure is within limits, I don't see how you could be exceeding terminal velocity?   This is just intuition speaking, but it SEEMS like some value of MaxQ set in Mechjeb would give you 'good enough' results...  when the alternative is hours of coding and another plugin in memory?  If not, I'd love to hear why my intuition is wrong so I can learn more about the subject!  

If you DO want to go the KRPC route - I've got a script that uses a PID to control thrust based on Dynamic Pressure pretty solidly.  It would be changing 3 or 4 lines of code to make it limit it based on any other parameter.  And while that's not a 15 second project if you're learning python at the same time...  I'd encourage you not to dismiss the fun you can have if you DO get into coding.

I view writing the software for my rockets to be just as important and fun as slapping the hardware pieces together!   For example - it's cool to have an abort routine watching the guidance numbers...  so if your rocket loses stability it could automatically chop power, lock SAS to retrograde, separate the crew capsule and deploy parachutes!

 

 

Edited July 1, 2017 by artwhaley

[linuxgurugamer]

2 hours ago, artwhaley said:

I absolutely am not volunteering @linuxgurugamer , one of the hardest working modder I know, for more work...  but...  if you take a look at this thread -  

If they DON'T have a part that responds to temperature, I suspect that it's something that might get added to the eventual todo list there if you explain exactly what you'd want it to do, in context of the other parts already in that mod?

That said...  and I'm not a fluid dynamics expert...   Isn't terminal velocity essentially a property OF dynamic pressure?   If the dynamic pressure is within limits, I don't see how you could be exceeding terminal velocity?   This is just intuition speaking, but it SEEMS like some value of MaxQ set in Mechjeb would give you 'good enough' results...  when the alternative is hours of coding and another plugin in memory?  If not, I'd love to hear why my intuition is wrong so I can learn more about the subject!  

If you DO want to go the KRPC route - I've got a script that uses a PID to control thrust based on Dynamic Pressure pretty solidly.  It would be changing 3 or 4 lines of code to make it limit it based on any other parameter.  And while that's not a 15 second project if you're learning python at the same time...  I'd encourage you not to dismiss the fun you can have if you DO get into coding.

I view writing the software for my rockets to be just as important and fun as slapping the hardware pieces together!   For example - it's cool to have an abort routine watching the guidance numbers...  so if your rocket loses stability it could automatically chop power, lock SAS to retrograde, separate the crew capsule and deploy parachutes!

 

 

The C# code for this isn't hard.  So if someone wants to make me a part model which is suitable to indicate temperature, I would be willing to add a temperature sensor.

On 6/29/2017 at 8:51 PM, Starlionblue said:

Good morning.

I'm seemingly obsessed with Eve ascents. Constantly trying to find the lightest weight vehicle.

I don't hand fly. Mostly I use MechJeb. However one of the problems I face is suboptimal thrust control. MechJeb allows setting of a dynamic pressure (Q) limitation on thrust, which is good for the initial ascent. However later I have a feeling that higher thrust is better due to an increase in terminal velocity.

I would like to be able to control thrust by matching vessel velocity (EAS) to estimated terminal velocity, both numbers found in Aero GUI from the debug menu. Another option would be thrust to be limited by a specific local Thrust to weight ratio. Say TWR reaches 2.0. At that point thrust stays at 2.0 and does not go higher.

Is there any add-on that allows setting thrust to follow various parameters like this? Alternatively, perhaps someone would like to take the challenge on?

Kind regards.

You may want to look at ConstantTWR, which I'll be releasing tomorrow. 

[Starlionblue]

Thanks for input. Much appreciated.

I can use dynamic pressure as a (manual) input for thrust control. However given rocket configuration changes during climb, the ideal number changes because terminal velocity is specific to the configuration. As side boosters are shed, terminal velocity seems to increase, and thrust should ideally increase.

Adjusting thrust on the fly by changing the Q limit number in MechJeb is finicky. I'll check if Smart Parts allows a staging event to trigger a change the thrust level. That would neatly deal with the problem.