When the above version of a Titan II is flown with MechJeb's PVG, using 80m/s turn start, 0.8 m/s turn speed, into a 161x225km orbit (Gemini 3's parameters), you separate at about 58.6km altitude whilst going about 2599 m/s, matching the planned for Gemini ascent profiles fairly well. At this point, the LR91 still has a 10% ignition penalty from Q, giving it absolutely terrible ignition reliablity.
I propose the ignition penalty tag for the LR91 be removed. If done so, during a typical Titan II ascent the LR91 can first be ignited without a Q-penalty at 42km going 1736 m/s, which gives a buffer of reasonable size for second stage ignition.
In January TestFlight received an update in which shielded engines also receive Q-penalties. https://github.com/KSP-RO/TestFlight/releases/tag/v2.8.0.0
This has lead to engines like the LR91 suffering from an approximately 7% ignition penalty at first stage cutoff when flown on a typical Titan II ascent profile, which calls for 1st stage separation at approximately 60km https://www.nasa.gov/wp-content/uploads/static/history/SP-4002/images/fig39.jpg
When the above version of a Titan II is flown with MechJeb's PVG, using 80m/s turn start, 0.8 m/s turn speed, into a 161x225km orbit (Gemini 3's parameters), you separate at about 58.6km altitude whilst going about 2599 m/s, matching the planned for Gemini ascent profiles fairly well. At this point, the LR91 still has a 10% ignition penalty from Q, giving it absolutely terrible ignition reliablity.
The main reason the LR91 suffers more from this problem than other engines is because it has the ignitionDynPresFailMultiplier = 0.1 tag, giving it a Q-penalty modifier of 0.15 at 5 kPa instead of at 50 kPa like other engines such as the AJ10 Early.
https://github.com/KSP-RO/RealismOverhaul/blob/8fc1266cea3007a29af98aac1fa63497f5638773/GameData/RealismOverhaul/Engine_Configs/LR91_Config.cfg#L255
This tag was added back in 2016. https://github.com/KSP-RO/RealismOverhaul/commit/6f3c0b8266b4549bf51fe3a54eaf3630de19afec The Titan I version of the engine does not have this tag, and there doesn't seem to be a specific explanation why the Titan II versions of the LR91 should have gotten it.
I propose the ignition penalty tag for the LR91 be removed. If done so, during a typical Titan II ascent the LR91 can first be ignited without a Q-penalty at 42km going 1736 m/s, which gives a buffer of reasonable size for second stage ignition.