dV-Calculations Incorrect with LH2 Engines

[StarStreak2109]

Hello all,

I am posting this here, because I have noticed some odd behaviour in conjunction with engines using liquid hydrogen (LH2), such as from the Cryo Engines pack by @Nertea or from BDB by @CobaltWolf. I am not posting the relevant threads, as I believe this is not a bug of these specific mods but rather something underlying, which makes these engines work.

The issue is that with any LH2 engine, that I have used so far, the dV calculation within the game seems incorrect. This becomes apparent, when using a mod such as MechJeb and/or KER, which calculate the dV per stage and for the whole ship.

Following please find some screenshots to illustrate what I am referring to: In the first screenshot, I am at the beginning of executing a manouver node of ~350m/s, I have 3,770m/s left in my rocket.

You would expect that after the burn, you'd be left with ~3,418m/s.

Instead, you end up with roughly twice the dV expended, with only 3,114m/s left in your rocket.

FWIW, I installed KER as well, to see if the error can be traced back to some error with MechJeb, but this does not seem to be the case, as the remaining dV is calculated almost to the same value (#floatingpointerror) in both mods. Now, I do believe there is only one way to calculate the dV of a rocket, hence the error must be somewhere else?!

I heard that others, @RocketPCGaming specifically, wondered about this inconsistency as well. As I have currently no idea, what might be causing this, I would like to stick some heads together to try and find the cause for this. As a start I would like to ask others, whether or not they have seen the same kind of behaviour and if so, was/is it similar to what I have seen (roughly twice the dV expended than anticipated) or was it a different kind of inconsistency?

Your feedback will be very welcome!

 

[4x4cheesecake]

4 hours ago, StarStreak2109 said:

As a start I would like to ask others, whether or not they have seen the same kind of behaviour and if so, was/is it similar to what I have seen (roughly twice the dV expended than anticipated) or was it a different kind of inconsistency?

Happens to me as well but with stock engines/fuel and AFAIK, the reason for this behaviour is somehow connected to the root part of the vessel.

Recently, my Cassini-Huygens probe I've build for one of the shuttle challenges, displayed a dV of ~2700m/s and just a single stage because the root part was on of the SRBs in the first stage. Re-rooting it to the probe core at the top brought back the correct values (1955m/s for the transfer stage and 322m/s for the probe itself)

Based on this observation, I assume that KER and MJ don't use the staging bar to separate the calculations for different stages but the part informations in the .craft file.

Any chance you got an unusual root part as well?

[CobaltWolf]

Wouldnt it be from the boiloff?

[StarStreak2109]

6 hours ago, 4x4cheesecake said:

Happens to me as well but with stock engines/fuel and AFAIK, the reason for this behaviour is somehow connected to the root part of the vessel.

Recently, my Cassini-Huygens probe I've build for one of the shuttle challenges, displayed a dV of ~2700m/s and just a single stage because the root part was on of the SRBs in the first stage. Re-rooting it to the probe core at the top brought back the correct values (1955m/s for the transfer stage and 322m/s for the probe itself)

Based on this observation, I assume that KER and MJ don't use the staging bar to separate the calculations for different stages but the part informations in the .craft file.

Any chance you got an unusual root part as well?

Hi,

interesting idea. I shall investigate that one!

2 hours ago, CobaltWolf said:

Wouldnt it be from the boiloff?

No, I have boiloff deactivated. Also in that short time span, 200% more fuel expenditure than projected at the beginning of the burn would be excessive, wouldn't you think?

[DoctorDavinci]

So not sure if it is related but BDAc AI air battles using craft built in previous versions of KSP use fuel faster in KSP 1.4.5 as compared to previous versions of KSP

I have craft I built going back a few years and I have found that on average the crafts in question use about a third or more fuel than they did in previous versions of KSP 

A plane I built specifically to fly for 15 minutes in a previous KSP version will only just make the 10 minute mark and then run out of fuel in KSP v1.4.x

Edited September 5, 2018 by DoctorDavinci

[Rocketology]

@StarStreak2109 I hadn't dug into this, just noticed it. I thought it may have been an error in KER as I was using SSTU which does things very different from stock/ other mods and KER miscalculates at times. That being said, have you done the rocket equation on a test vehicle and compared it to it's actual performance vs. KER/ MJ?

And as Cobaltwolf stated above, have to account for boiloff which varies between different mods as to the rate and other factors such as tank type/ EC usage.

. 

 

Edited September 13, 2018 by RocketPCGaming
added content

[StarStreak2109]

Hey,

actually, I hadn't had much chance to dig into it deeper due to real life constraints... I have not thought of doing the math actually. My findings / prerequisites are still:

• KER and MJ deliver roughly the same numbers concerning dV for both LFO as well as LH2 engines.
• When using LH2 engines, the actual amount of dV spent during an in-space manouver is twice as high as stated by the manouver node system.
• Boil-off is deactivated all the time, so it should not factor in the calculations.

I still have to try whether the root part has anything to do with it...

[Ultimate Steve]

You know, I've had something similar happen... Don't know what is causing it, but I ended up with more Delta-V than what KER said. MJ consistently showed more (No, I wasn't in Aero mode) and as a result my 100t lifter can lift something like 140t, which shrunk my lunar architecture form one launch to two.