Procedural SRB thrust reduced to 719 kN at launch

[HebaruSan]

If a procedural SRB is configured for a thrust in excess of 719 kN, the thrust will be reduced to 719 kN at launch. If this is an intentional limit, it should be reflected in the editor in the VAB. If it's unintentional, it should be removed.

Screenshots

SRB with thrust 7 MN in VAB:

Same craft plus launch clamps on launch pad:

Steps to reproduce

1. Go to VAB
2. Place a command pod
3. Place a procedural SRB under it
4. Make the SRB big, I used a curved cone with a 2.5m top, 3.75m bottom, and 24m length
5. Set the thrust to the maximum allowed, in my case about 7 MN (or "kkN" as it says)
6. If installed, KER will show a TWR in line with the configured thrust
7. Click Launch
8. Notice that the engine bell is a lot smaller on the pad than it was in the VAB
9. Press space
10. Notice that the actual thrust is only 719 kN

Craft file: Auto-Saved Ship.zip

Mods installed

• Procedural Parts (Starwaster/master branch current version)
• Porkjet part overhauls (no parts from this mod used in test)
• Module Manager 2.8.0

[Starwaster]

This problem has to have been around for awhile but it doesn't happen with RF because (apparently) PSRB uses different methodology if there are no ModuleEngineConfigs present. I guess most of us using PP are also using RF?

I have a general idea of where in the code this is happening but I'll have to poke at it some more before I understand WHY.

Personally though, I've never liked how PSRB calculates thrust. It's a little backwards compared with how SRBs work IRL, but I was hesitant to change things because it might break existing craft. Maybe if I had then I would have noticed this other thrust issue.

[BloodyRain2k]

Since this would most likely result in breaking existing crafts maybe use the chance, since things might get broken anyways, to add thrust profiles?

Dunno if RF has that, I'm one of the few weird ones that play without RF and just want to not be limited by stupid stock tank sizes which PP does a great job at.
But SRBs are sadly almost useless because of the thrust limit, and the bigger the SRBs get the bigger the variation in TWR to the point where I had some rather big S2 SRBs with max thrust of the mentioned 719kN on the pad and at takeoff my craft had a TWR of 1.8 and shortly before they burned out (over 1min later) it was at 4.5 which is pretty excessive if you're still in the atmosphere.

Honestly no idea though how that would be done, both code and UI wise.
Maybe just a toggle for them to throttle down proportionally with loss in mass so the overall TWR stays consistent?
Would be enough for me but well, I'm not the one coding them and not the only one using them : /

Just as a suggestion.

Oh and while at SRBs: I once tried to use them as replacement for the seperatrons and well, it didn't work as well as I hoped, probably because those are for their size excessively overpowered in thrust and capacity.
What really surprised me though was that PSRBs do not scale their thrust plume so I had rather tiny PSRBs with minimum thrust of 10kN and giant thrust plumes. Scared the hell out of me because I thought things were exploding all over the craft. So maybe add this to the list of eventual fixing?

[HebaruSan]

@BloodyRain2k , I would recommend creating separate issues for each thing you want fixed. It's not too easy to track bug reports that are posted as one sentence in the middle of a long comment replying to a different issue.

https://github.com/Starwaster/ProceduralParts/issues

[BloodyRain2k]

Sorry, bad habit of mine, I tend to squish things together when I can, even when it doesn't help...
Made one now.

[hippomormor]

I can confirm this also happening with the RealFuels SRB's

[hippomormor]

Changing line 571 in ProceduralSRB.cs fixes this issue when RealFuels is installed, maybe also for non-RF:
- private float attachedEndSize = 1.25f;
+ private float attachedEndSize = float.PositiveInfinity;

EDIT: Just tested without RealFuels - It works with the craft provided by @BloodyRain2k

[Starwaster]

It seems likely to me that the only way that would work is if it's forcing a boolean check to pass that would otherwise file, such as in OnPartAttachNodeSizeChanged. So the problem is likely that UpdateMaxThrust is not being called as a result of the failed check.

It would be preferable therefore to discard the node size check and ensure UpdateMaxThrust executes rather than throw Infinity into the mix and run the risk that it ends up in a vital math operation. (not possible now, but that's a good way to end up with regressions later)

[Starwaster]

So it looks like the reason the infinity thing works is because attachedEndSize doesn't get updated in flight mode and maxThrust is affected (gets capped at 1.25² * 500). It IS updated in the editor which is why there's a mismatch between editor and flight. MOSTLY doesn't affect RF though it might not be as immune as I thought.

I might do the infinity thing as an temporary solution because I don't think I have the time to really do this properly, which would be to revamp how thrust is calculated and persisted. What I'd LIKE to do is change the actual calculation to be more like real SRBs, which is that length would affect thrust and diameter would affect duration. But then I'd also want to do it in a manner which doesn't affect any existing craft 'in the field'. I don't know if that will ever really happen though.

[wb99999999]

Having a more realistic calculation for P-SRBs would be extremely desirable for me personally. I would break my craft file for this any day of the year.

[Nomeo]

Just tell me how to fix it , it happens to me too

[Nomeo]

my mind is a nut , I need simplicity

[HebaruSan]

@Nomeo, there is unfortunately no known fix yet. Fixing it yourself would require debugging the mod's source code.

[hippomormor]

@Nomeo, if you're using KSP 1.2.2 I could send you my version?

[benguinsburg]

@hippomormor I'd love to see your fix as well