SSU Development thread (4.0 to 5.0) [DEVELOPMENT HALTED DUE TIME REQUIREMENTS!]

[GLS]

For the last couple days I have been trying to figure out why the GOX vents through MPS intermediate seal turbopump drain lines doesn't want to line up. I've checked everything. Everything seems to check out. Is there some logic I'm missing in the set up? The actual coordinates is with the nozzles in the installed NULL positions which I got directly from AC3D.

https://www.dropbox.com/s/24l5l3wrzhl4uh8/Misaligned_GOX_vents.jpg?dl=0

I've seen that before, and figured why that wasn't working... but I'm not remembering all the details. :facepalm: I think it has to do with the initial position of the engines.... for some reason I didn't, or more likely couldn't fix it. :shrug:

 

[DaveS]

I've seen that before, and figured why that wasn't working... but I'm not remembering all the details. :facepalm: I think it has to do with the initial position of the engines.... for some reason I didn't, or more likely couldn't fix it. :shrug:

So should I just leave things as-is for the moment?

 

[GLS]

So should I just leave things as-is for the moment?

Yes, when I'm more "free" from TAEM guidance I'll take a look again.

 

[DaveS]

I'm finishing up the orbiter and I got a problem. I need to tilt the orbiter when it is attached to the ET to get the proper alignment between the forward and aft attachment points. Where in the code can I change the direction and rotation of the orbiter attachment point to the ET? I got the vectors ready but I need to know where they go.

 

[GLS]

I'm finishing up the orbiter and I got a problem. I need to tilt the orbiter when it is attached to the ET to get the proper alignment between the forward and aft attachment points. Where in the code can I change the direction and rotation of the orbiter attachment point to the ET? I got the vectors ready but I need to know where they go.

I'm not sure that will be possible to do, at least not in a couple of minutes (or more likely hours), as the engines are coded to point to the c.g., which is calculated assuming the ET and OV are aligned, yada, yada, yada...
I'm currently putting the final touches on TAEM and am already thinking Entry guidance, so only after that will I have my brain "free" for other big things.

For now I suggest splitting the difference between the fwd and aft attach points, leaving the orientation as is, and updating ticket 186 to reflect this remaining work. :shrug:

 

[DaveS]

I'm not sure that will be possible to do, at least not in a couple of minutes (or more likely hours), as the engines are coded to point to the c.g., which is calculated assuming the ET and OV are aligned, yada, yada, yada...
I'm currently putting the final touches on TAEM and am already thinking Entry guidance, so only after that will I have my brain "free" for other big things.

For now I suggest splitting the difference between the fwd and aft attach points, leaving the orientation as is, and updating ticket 186 to reflect this remaining work. :shrug:

One thing that has to be done along side ticket 186 is the animation of the umbilical disconnect plates so that they mate properly with the ones on the ET.

 

[Urwumpe]

Also, if the orbiter is suddenly angled relative to ET and SRBs, we might need to check if the SRB thrust vectors and hold down posts are still implemented properly.

 

[DaveS]

Also, if the orbiter is suddenly angled relative to ET and SRBs, we might need to check if the SRB thrust vectors and hold down posts are still implemented properly.

AFAIK, The ET/SRBs have a neutral pure vertical angle so there shouldn't be any changes there. The reason for the angled orbiter is that the forward attachment point (the bipod struts/yoke) are taller than the aft ball/socket joints. The bipod has height of 56.341" (1.431 m) while the aft ball/socket joints have a height of 41.259" (1.048 m). As you can see, the aft sits lower than the forward, creating a disconnect between the two if the angle was flat and level.


For reference, the separation planes for the bipod and ball/socket joints are as follows:
X_o283.841 (bipod)
X_o267.556 (aft ball/socket).

 

[GLS]

AFAIK, The ET/SRBs have a neutral pure vertical angle so there shouldn't be any changes there. The reason for the angled orbiter is that the forward attachment point (the bipod struts/yoke) are taller than the aft ball/socket joints. The bipod has height of 56.341" (1.431 m) while the aft ball/socket joints have a height of 41.259" (1.048 m). As you can see, the aft sits lower than the forward, creating a disconnect between the two if the angle was flat and level.


For reference, the separation planes for the bipod and ball/socket joints are as follows:
X_o283.841 (bipod)
X_o267.556 (aft ball/socket).

... and the bipods couldn't be made to fit the OV, because...??? The more I think about a "titled OV", the more it doesn't make sense...

 

[DaveS]

... and the bipods couldn't be made to fit the OV, because...??? The more I think about a "titled OV", the more it doesn't make sense...

It is angled. Not by much, just 0.1°. They do angle the orbiter when it's mated to everything else, including the OTS and the SCA. It's the same for both, 2°s. For the ALTs, the angle was raised to 5°s by the use of a taller bipod assembly on the SCA.

---------- Post added at 09:35 PM ---------- Previous post was at 09:24 PM ----------

You might be interested in this document: Space Shuttle Separation Mechanisms

 

[GLS]

It is angled. Not by much, just 0.1°. They do angle the orbiter when it's mated to everything else, including the OTS and the SCA. It's the same for both, 2°s. For the ALTs, the angle was raised to 5°s by the use of a taller bipod assembly on the SCA.

---------- Post added at 09:35 PM ---------- Previous post was at 09:24 PM ----------

You might be interested in this document: Space Shuttle Separation Mechanisms

Yes, it is visibly nose-up in the SCA, but the requirements are totally different: the OV sits passively in the SCA, versus having to point engines and manage IMU and RGA data during launch. I'm not saying it is impossible, just that it adds extra complexity, so there must be a very good reason to do so. Where did those 0.1º come from?

BTW: I read the document "in the diagonal" and it doesn't seem to add anything to this discussion.

 

[Urwumpe]

Well, remember we are using the orbiter body coordinate system and all others (ET, SRB) are relative to the orbiter, even on the pad. Its no really big problem for the stack and can even be ignored for SCA and others.

But any such alignment should be handled with care, because a lot depends on it.

 

[DaveS]

This is with the orbiter perfectly level on the ET with the attachment point being right for the bipod: https://www.dropbox.com/s/yxoc5ghrgs5ijrj/AC3D_AFT_OV_ET.jpg?dl=0


As you can see we got interference with the aft diagonal thrust struts when in reality the orbiter sits a bit above them and the cross-beam. This photos shows it quite well: https://www.flickr.com/photos/apacheman/5801795963/ as does this schematic of the aft ET hardware: https://www.dropbox.com/s/usklpunf1nl0lj5/ET_OV_Aft_Attachments.jpg?dl=0

And this is with a 0.2° angle: https://www.dropbox.com/s/zlbcrhfp1feeeta/AC3D_AFT_OV_ET_02_angle.jpg?dl=0

---------- Post added at 12:58 PM ---------- Previous post was at 12:33 PM ----------

The 0.2° angle also negates the angle of the umbilical disconnect plates on the orbiter which are also angled due to the bottom shape of the orbiter aft engine compartment.

 

[GLS]

This is with the orbiter perfectly level on the ET with the attachment point being right for the bipod: https://www.dropbox.com/s/yxoc5ghrgs5ijrj/AC3D_AFT_OV_ET.jpg?dl=0


As you can see we got interference with the aft diagonal thrust struts when in reality the orbiter sits a bit above them and the cross-beam. This photos shows it quite well: https://www.flickr.com/photos/apacheman/5801795963/ as does this schematic of the aft ET hardware: https://www.dropbox.com/s/usklpunf1nl0lj5/ET_OV_Aft_Attachments.jpg?dl=0

And this is with a 0.2° angle: https://www.dropbox.com/s/zlbcrhfp1feeeta/AC3D_AFT_OV_ET_02_angle.jpg?dl=0

---------- Post added at 12:58 PM ---------- Previous post was at 12:33 PM ----------

The 0.2° angle also negates the angle of the umbilical disconnect plates on the orbiter which are also angled due to the bottom shape of the orbiter aft engine compartment.

In other words, there is not "official" or direct data that says the OV is angled*, only measurements that work better with an angle... well, in this particular case that is not going to be enough to change code, as the implications of that angle not being 0 are just too big.
IMO, the likelihood of one of those measurements having an error (in value, conversion or just tolerance stack-up) is higher than the OV being at an angle, when we consider the GNC issues that would bring (that again, are not insurmountable).

*) in fact, doesn't the SODB (or the large diagrams books) have some diagrams with the axis of the 4 stack components, and their axis matching each other?

 

[DaveS]

In other words, there is not "official" or direct data that says the OV is angled*, only measurements that work better with an angle... well, in this particular case that is not going to be enough to change code, as the implications of that angle not being 0 are just too big.
IMO, the likelihood of one of those measurements having an error (in value, conversion or just tolerance stack-up) is higher than the OV being at an angle, when we consider the GNC issues that would bring (that again, are not insurmountable).

*) in fact, doesn't the SODB (or the large diagrams books) have some diagrams with the axis of the 4 stack components, and their axis matching each other?

Well, the angle is very minute and documents with graphics have a very hard time showing very minute details such as sub-degree angles. I just drew a line 0.2° line myself in Paint.NET and it is not even detectable. I can ask someone who should definitely know, a retired NASA-KSC Shuttle Mechanical Systems (group callsign MEQ) engineer who worked alot of orbiter/ET mates (and few demates) right up to the very end. He should know if the the orbiter had an angle when mated.

 

[GLS]

Well, the angle is very minute and documents with graphics have a very hard time showing very minute details such as sub-degree angles. I just drew a line 0.2° line myself in Paint.NET and it is not even detectable. I can ask someone who should definitely know, a retired NASA-KSC Shuttle Mechanical Systems (group callsign MEQ) engineer who worked alot of orbiter/ET mates (and few demates) right up to the very end. He should know if the the orbiter had an angle when mated.

Please ask, as it's much better to hear it from the "horse's mouth" than us here trying to guess.
I'm sure there were some tolerances (like 0.0º +/-0.1º), but the "normal" angle is what we need.

 

[DaveS]

I have sent the message with the question. I'll relay anything worthwhile.

---------- Post added 09-08-18 at 12:34 AM ---------- Previous post was 09-07-18 at 06:05 PM ----------

Can someone verify the math in this post on the NASASpaceflight.com forum by Fequalsma: https://forum.nasaspaceflight.com/index.php?topic=17437.msg1853722#msg1853722

1.1°s sounds a bit much.

 

[GLS]

Can someone verify the math in this post on the NASASpaceflight.com forum by Fequalsma: https://forum.nasaspaceflight.com/index.php?topic=17437.msg1853722#msg1853722

1.1°s sounds a bit much.

The math is OK. It assumes we want to calculate the angle between the attachments, which is not what we need... :uhh:
We need the angle between the axis of the OV and ET.

 

[DaveS]

The math is OK. It assumes we want to calculate the angle between the attachments, which is not what we need... :uhh:
We need the angle between the axis of the OV and ET.

Can't that be calculated from the angle between the attachments? They're where they are, that doesn't change. And the orbiter axis doesn't change, it's still level with everything, it's just that the orbiter has a slight positive AoA when mated to the ET. Obviously Enterprise never had any issues with the 6° AoA used for the ALTs.
And the shuttle GNC could easily sense the slight AoA as the reference plane never changed.

 

[GLS]

Can't that be calculated from the angle between the attachments?

That assumes that the attachment points in the OV are leveled, which they aren't.

And the orbiter axis doesn't change, it's still level with everything, it's just that the orbiter has a slight positive AoA when mated to the ET.

These 2 sentences contradict each other. If the axis are leveled, there is no angle. :shrug:

Obviously Enterprise never had any issues with the 6° AoA used for the ALTs.
And the shuttle GNC could easily sense the slight AoA as the reference plane never changed.

The Enterprise wasn't controlling the mated flight...
I conceed that there may well be an angle in there for aero reasons, but I want to confirm it really exists before I change tons of code. "3D models don't fit" isn't enough evidence for this, especially when there isn't a single bit of info pointing to some angle existing in there, plus there is info saying the axis match, plus it would make GNC harder.

---------- Post added at 11:40 AM ---------- Previous post was at 12:23 AM ----------

I just committed a scenario to run TAEM at Edwards. It should allow all TAEM and A/L trajectory options to be "exercised". FCS is (still) "wonky", so although most guidance references are correct*, they might not be followed perfectly.
Tons of things are still missing from the displays, others need fine-tuning. I'll work them after Entry guidance.


*) Some parameters are "stock" so the whole thing might not fit perfectly, but it gets to the runway. Also, TAEM has several requirements, so you might see e.g. altitude diverge for a while, but that is to correct Qbar or E/W. PRFNL roll has a small issue... I'll track that down later.