MPS development

[DaveS]

New SSME

Here's a new SSME I have been making for the past few days. Is this something that is wanted on the official orbiter mesh?

 

[Urwumpe]

How many polys do three of them have?

 

[DaveS]

How many polys do three of them have?

About 21k(7k each).

 

[Urwumpe]

About 21k(7k each).

Hmm, 7K would have been already painful for the mesh... 21k is far out.

And I don't feel like I can get started on doing LOD calculations on the meshes.

 

[DaveS]

Hmm, 7K would have been already painful for the mesh... 21k is far out.

And I don't feel like I can get started on doing LOD calculations on the meshes.

Hmmm, how about we do what is already done on the default Atlantis and Kulch's Energia? We go with a configurator, where the user could choose between using the high-res meshes or the low-res meshes?

That way it shouldn't be too painful for those with lower-end machines. And animations shouldn't be impared as we're using meshnames and not mesh-group numbers to call the correct components.

How about this idea?

 

[Woo482]

if I am correct you can do most of the details with texture's and if orbiter supports it parallax occlusion maps can add 3d details to a 2d texture

EDIT: 2d texture with parallax occlusion mapping

 

[Donamy]

Very nice work David !! I'd like to take a look at those meshes, and see if I can reduce them some. Or, do you think they are reduced enough already?

 

[DaveS]

Very nice work David !! I'd like to take a look at those meshes, and see if I can reduce them some. Or, do you think they are reduced enough already?

Well, see if you can't optimize it some. Each engine is identical. However, the #2(Left) is clocked 90°s counter-clockwise. The mesh is clocked for the #1(center) and #3(right) positions.

 

[Donamy]

Try this, first back up your mesh.

 

[GLS]

I did not yet change this one, as long as the SSME code of GLS was still untested. Wanted to keep the chance to disable the new code if it does not yet work.

Also, it works when the Shuttle is not attached to the pad.

My code is still not active...

 

[Urwumpe]

My code is still not active...

You just had to remove the comments from the lines in the Atlantis constructor to enable it.

I just edit some lines in it, so it does use the existing SSME thrusters.

PS: How do you want to set the thrust level?

0.0 - 1.0
or

0 to 109% ?

 

[Urwumpe]

OK, SSME code is enabled now and should be called, though it might interfere with the normal launch behavior.

I think I will move the MPS SOP on top of the list of software to be implemented in the GPC, so we can quickly have control over the SSMEs.

 

[GLS]

You just had to remove the comments from the lines in the Atlantis constructor to enable it.

I just edit some lines in it, so it does use the existing SSME thrusters.

PS: How do you want to set the thrust level?

0.0 - 1.0
or

0 to 109% ?

It's 65% (or 67%) to 100% (or 109% or whatever the engine allows...) (67-109 for Block II)

BTW: You should slow down just a little because I've done some changes to the code... and I haven't uploaded them yet.....:sorry:
But if you use the EIU there shouldn't be any impact... I think...
I'll take a look at your changes and upload a working version tomorrow...

 

[Urwumpe]

It's 65% (or 67%) to 100% (or 109% or whatever the engine allows...) (67-109 for Block II)

BTW: You should slow down just a little because I've done some changes to the code... and I haven't uploaded them yet.....:sorry:
But if you use the EIU there shouldn't be any impact... I think...
I'll take a look at your changes and upload a working version tomorrow...

Just merge your changes into them, I did not so much, I just replaced the many direct reference to the thruster handle of the SSMEs by methods of the Atlantis class. No EIU yet, but I started refining the Shuttle Bus elements into something simple, so the EIU could be

I had a unscheduled CTD in Orbit, will look what caused it.

For implementing the hydraulic and systems, what about making a HydSystem class, which handles pressure and return lines?

The rest seems to be a typical parallel topology, so all hydraulic actuators get the same input pressure and the return is the result of the actuator work.

If we can formulate the behavior into functions of the type

dP = f(P, ...)

we should be able to have these be calculated numerically in real time.


So, we would finally have:

APU<->HydSystem<->Actuator

As each APU supplies one Hydraulic system, it might even make sense to unify both classes into one.

 

[DaveS]

Hmmm, speaking of MPS and performance: I think we need to take a look at lift-off. The vehicle leaves the MLP way too fast right now compared to the real thing. I don't know if it's due to lack of mass in the payload bay or that we have too high performance figures, but it needs some attention.

 

[Urwumpe]

Hmmm, speaking of MPS and performance: I think we need to take a look at lift-off. The vehicle leaves the MLP way too fast right now compared to the real thing. I don't know if it's due to lack of mass in the payload bay or that we have too high performance figures, but it needs some attention.

How much too fast? I think I remember there had been some performance tables around, could be used for that task.

I remember that our vertical speed was quickly higher as stated in the checklist, but I thought it is because of the bad approximation and small differences in the pitch table.

 

[DaveS]

How much too fast? I think I remember there had been some performance tables around, could be used for that task.

SSU tower clear: T+3 sec.
STS-114 tower clear: T+5 sec

 

[Urwumpe]

SSU tower clear: T+3 sec.
STS-114 tower clear: T+5 sec

Hmm... Can you check if we start from the pad with zero initial speed?

 

[DaveS]

Hmm... Can you check if we start from the pad with zero initial speed?

Heck no! It's instantly from 0 to 23.3 m/s² VSpeed! It should be alot lower than that!

 

[Urwumpe]

Heck no! It's instantly from 0 to 23.3 m/s² VSpeed! It should be alot lower than that!

OK, then it is caused by the orbiter bug of using thrusters while being attached. I remember I found a fix for it for the SubOrb project, could look into the old sources.