[1.0.2] NovaPunch 2.09. - May 6th - 1.0 Compatibility Update

[Tiberion]

Are you bumping the weight of all your payload too?

Btw if you didn\'t know, there is a separate zip on the front page now with just the engine and tank configs in it, so people can use the 0.3 (or even stock SIDR/SP numbers you\'re welcome) if you need an easy way to duplicate them for testing.

[Commander Zoom]

IMO, the first set of modified numbers - normalized to match the ratios of the stock parts better than SP\'s original approximations - were just fine. If there\'s a problem with the capabilities of parts with those numbers, it\'s up to HarvesteR to rebalance the stock parts, again IMO.

[Thunderbird]

i mostly agree, most of the part packs have their values derived from stock,atleast approximately, and id prefer if it would stay that way so they are compatible, within boundaries ofcourse

if there is something unbalanced though,that doesnt mean keep it that way... balance the hell out of it, just keep it within the compatibility range

either way,keep up the good work Tiberion, :thumbup:

[Commander Zoom]

The thing is, it sounds like in this latest attempt, Tiberion decided it should be harder and more 'realistic', and cranked the handicap way up over even the stock parts... and found out, oops, that was too far and the fun little rocket game no longer is.

IMO, this project - which started as nothing more than an attempt to collect and keep current two widely-used, but no longer supported, early parts packs - is beginning to suffer from a bad case of feature and/or mission creep. If the original goal has been met, then stick a fork in it and start a new project.

[Andras]

ONE ISSUE IS THAT OF VOLUME. THE (sorry, had it on precision mode!) large 3m tank would have a volume equivalent of about 33000 fuel. That\'s bananas though, weighing about 150 using strict ratios.

fuel/volume though is easy to rationalize, different fuels, etc

Another is net weight, how important do you/we consider the dry mass of the tank to reduce the fuel count? The stock LFT looses 10% to dry mass, and it has an efficiency of 222.22 fuel per of net weight

Say a 3m tank needs 30% of it\'s fuel mass for support, and a 2m tank needs 20% (1m tanks 10%). That would help reduce the both the fuel counts, and the lifting capacity of the motors. It also helps slow down the high acceleration when the tank begins to run dry.

Forex, a 15000 count 3m tank has a fuel mass of 67.5, and a dry wt of 20.25, for a total wt of 87.75

To work that backwards, a gross wt 60 tank 3m in dia, could have a dry wt of 13.8, and a net fuel mass of 46.2, or 10,266 fuel.

BRB, have another test flight to run.

[Tiberion]

Um? No feature creep.

I wasn\'t aiming for more realistic as much as balanced, the way it was made it dumb and it was NOT balanced against stock since one tank of equal weight held 2x as much fuel - thats because the IKAO calculator is based on 'real' ratios.

Yesterday was an attempt to make the fuel usage reasonable so that one 3m stage wouldn\'t get you to Mars (I had the full 1.75m stage and the upper 1m stage AND the lander stage all in Munar orbit with full tanks, thats just not reasonable for such a conservative rocket stack)

Anyway, I\'m going back to the drawing board for 0.5 and running the volume numbers myself - the tanks WILL be standardized against the stock tank.

Then I\'ll figure some reasonable thrust levels to lift those tanks, and then the fuel usage will be balanced to stock ratios (25 to 1, thrust to fuel)

And do remember to keep that 'ALPHA' tag at the end of my downloads in mind - I\'m still testing values here to make he pack work as a unit. Just cramming the parts in a folder and releasing it wouldn\'t work, the 2 packs used different ratios.

Andras: I think your Cylinder volume measurements might be off. The stock tank is 1x1.75m and the big 3m tank is 3x6

Using Pi*r^2*h to figure volume, the stock fuel of 500 scales up to 10250 fuel in the 3x6 tank, for 50mass

I\'m gonna hammer out the numbers and ignore scaling dry weight against cylinder size until the next version so I can get some baseline numbers.

Edit: here is the volume -> fuel figures after my calculations. Red is SIDR/SP numbers, Green was the 0.1 NovaPunch numbers, and blue are the new one\'s as of today.

If you\'re refreshing the spreadsheet from the 1st post, the engine values aren\'t updated to reflect this yet, that is next.

[Andras]

Stack up 4 tanks next to the 3m large tank in the VAB and they are the same height, but that\'s still only 18,000 fuel, not sure how I got 33,000.

The 2m long and 2m short together are the same height also, so that\'s what, about 2650/5300 respectively? I guessed 2500/5000

3m short that same height as the 2m long, so 12,000?

The 3m quad had good results with a thrust of 2,000, burn of 100, dry mass of 10.

We really need a 3m \'second stage\' motor that\'s not as big as the quad. I suppose the 1m adapter plates will do.

Looking forward to what you come up with.

[CAPFlyer]

To reply to your Vernier thing - I would rather it be an LFE, that\'s what they are in real life - constant burning motors that add a tiny bit of thrust, but are mainly used to constantly direct the vehicle. To turn them into big, powerful, RCS thrusters would be to defeat their purpose. Additionally, the real vernier thrusters usually use the same fuel supply as the primary engine, so that adds to the need for them to be LFE\'s in KSP.

[Tiberion]

Yeah I decided to keep them as LFEs - trvtannenberg already made the RCS version and there\'s no need to duplicate them, and the LFE version adds some functionality.

Mainly LFE seems better to be because it can add control authority without the need for ALL RCS to be toggled on. If you could turn them on without also using thejets way up on the lander, it would be better, but we cannot, yet.

I have the new numbers plugged in. Once I get back from dinner I\'ll playtest them, but you can look them over in the spreadsheet - the blue numbers are now current.

https://docs.google.com/spreadsheet/pub?hl=en_US&hl=en_US&key=0Ah0q6AoAI_W6dE5fUHo3NkhEUG9CODYzTFpCRTU2OWc&output=html

[sarkun]

Andras: I think your Cylinder volume measurements might be off. The stock tank is 1x1.75m and the big 3m tank is 3x6

The stock 1m tank is 1.5 meter high I think. (1.515 if u take it\'s top and bottom node values).

[Tiberion]

eh, I guess it is.

But you made me discover my math error, so the big 3m tank is 18000, which at 200:1 is 90 mass units. Whee.

Fixing my math and testing. I already had to bump up the collision tolerance on every single part due to the weight of the earlier 12k numbers... sigh.

[sarkun]

No worries - when I was modeling my first parts, I assumed that it was 1m high. I loaded the part into KSP, sighed and made another... this time assuming it was 2m... THEN I checked

[Guest Flixxbeatz]

Vernier pod converted from SAS module to radially-attached liquid engine (thrust 15, gimbal 10degrees)

[Tiberion]

Well this isn\'t going to work.

The 3x6m tank is 90mass units, and trying to load my rocket with those values breaks connections between the decouplers and engines, or engines and tanks. Its even worse than K&W. I simply won\'t tolerate having to use struts on every basic connection, just to reach stock 'scaled' properly.

So as an attempt to solve this, I am going to scale the tanks right, and then half all values, so the big tank will be 9000 fuel and 45mass

Since the Fuel:Weight ratio will be the same, and the engines are already light to reward staging, the older thrust values and the stock fuel usage should work okay as a starting point.

We\'ll see.. I still hope to get a new version out for people to try, I have a good number of bug fixes in as well.

[Andras]

I think we/you could easily knock the numbers on the 3m tanks back to 12000/8000, mainly to keep the weight numbers in check, and so we don\'t need ridiculously high numbers on the 1st stage engines. The dry weight of the tanks needs to go up, to at least 10% of the loaded weight, and I\'d prefer it to be a little higher. Just put a line in the description of them using a lower density fuel or increasing structure requirements as tanks get bigger, these guys are Kerbals, not German rocket scientists.

The 2m tanks could go to 2250/4500 or 2500/5000.

The long 1.75m tank is the joker, it\'s going to have a lot of capacity no matter what is done to it. At the same time you want to have a reason to use 2m tanks over the 1.75 tanks, arbitrarily set the long 1.75=the long 2m, and the mid sized 1.75= short 2m? shortest 1.75= half that?

ETA- I see your post you ninja you. I did ok with a 3m tank that weighed 60 with 15000 fuel.

[Tiberion]

Hmm.. how about this:

Due to tank construction, the 3m tank only hold 75% as much internally as a 1m tank would. The 2m and 1.75m are 90%

[Tiberion]

Well after taking a mind clearing break, and then slogging through the chore of adding breakingForce and breakingTorque entries to all relevant parts, I did manage to get a roughly working version going.

I\'m gonna post the zip has an X release, just so those interested can help me tweak the new numbers. I\'m not updating the 1st post til I get some issues ironed out.

There still seems to be an issue with the vernier doing damage to the engine and causing lag and freeze-ups, I\'l investigate that tomorrow.

Mainly I\'d like some feedback on how the engines needed adjusted in relation to the new weights - I made some rough guesses and tested, but I don\'t think its right yet. I may have bumped the 1.75m engines too far.

Here\'s the changelog:

NovaPunch v0.5alpha (X release)

-fixed vernier pod effects lag, increased thrust slightly

-added fuelflow = true to all non-decoupling adapters and RCS tanks that were missing it

-GG-M4 has correct node placement and name, it should once again work with old save files

Removed duplicate part '1m_1.75m_short_adapter' since an identical 'fueltank' version existed.

-added NP_landerleg - the model is the 'old' leg from Nova\'s SE edition

-increased crashTolerance, breakingForce and Torque on tanks, engines, parts that were set too low to tolerate the NP tank mass increase - if you find one thats still snapping on the pad, let me know

-Did a major re-balance pass on tanks and engines - the original metric I used probed problematic, so I started from stock and worked up. This is very much a test version of these numbers, expect issues - use the Alt-config zip on the front page to grab the previous versions 'stock' configs if you need them.

For testing, you can refer to the spreadsheet to compare old numbers - and if you absolutely need a bugfix but don\'t care to test the numbers, there is an Alternate Config zip with the old numbers, because.. well.. I care.

Sleep time, give me some feedback pls.

http://dl.dropbox.com/u/46506740/KSP/NovaPunch0_5X.zip

http://dl.dropbox.com/u/46506740/KSP/NP_Alt_Configs0_5X.zip

[Andras]

Too late for me now, but I\'ll check it tomorrow. I\'d like to request a bottom attach node for the 3m Omni Bertha in the official release. I added one at -0.38 and it seems to work with the short 3m shroud decoupler (-0.5 might be better). That lets us use it as a second stage engine instead of the huge quads (kuads? lol).

[Tiberion]

Yeah I can do that, I\'m surprised it fits since its slightly bigger than 3m.

I would like to replace it with another model since that engine really belongs with the N1 models. We\'ll see if I have the time.

Anyone tested this yet? Or does no one really care at this point?

[Andras]

I\'VE BEEN (dammit) flying test shots.

What I like so far:

Quad- mass 8, 3200 thrust, 160 burn, vectorable

(It\'s a 4x800 innit?)

Omni- mass 4.5 thrust 1800, 120 burn, vectorable

(many small motors not as efficient?)

2nd stage Burns out before orbit is established

No changes to your fuel tanks

I\'m trying to balance an Apollo clone built of out of these parts in the pack

3m large tank and Quad

3m short tank and Omni w short 3m decoupler

2m short tank and a twin bertha

1m lander and Kerbin return

I have a 3mx1m RCS tank above each 3m stage, and 5 rings of four RCS thrusters

I\'d like a regular SAS unit texture matching the restyled Yawmaster bits. I hate hate hate ASAS units. The constant thrust vectoring and RCS usage makes me bananas and my rockets bendy.

Test 1 was 100% throttle, Test 2 @ 2/3rds, I didn\'t run test 2 out to the Mun

[Tiberion]

So do you like the upscaled weight? I was getting joint wobble constantly, even with the boosted breakingforces, and I wasn\'t happy with how the rocket controlled in the lower atmosphere.

I decided to tes out how your suggestion to roll back to the 0.3 numbers and then work up a little to counter the silliness of the tank scaling - its working better.

The 3m tanks are 10000 and 6600 with 40 and 25 mass. That didn\'t require the boost to thrust on the engines so much either, which solved the wobblyness.

So we can either continue to work on polishing the 'stock' scaled values which result in the high-mass tanks, or I can stay closer to the 'relaxed' balance of 0.3 and lower, and just tweak the tanks to not be so laughably fuel-filled.

Its really about what the people who are likely to use the pack desire; 'realism' or 'flexibility'

[Straight Chillen]

So do you like the upscaled weight? I was getting joint wobble constantly, even with the boosted breakingforces, and I wasn\'t happy with how the rocket controlled in the lower atmosphere.

I decided to tes out how your suggestion to roll back to the 0.3 numbers and then work up a little to counter the silliness of the tank scaling - its working better.

The 3m tanks are 10000 and 6600 with 40 and 25 mass. That didn\'t require the boost to thrust on the engines so much either, which solved the wobblyness.

So we can either continue to work on polishing the 'stock' scaled values which result in the high-mass tanks, or I can stay closer to the 'relaxed' balance of 0.3 and lower, and just tweak the tanks to not be so laughably fuel-filled.

Its really about what the people who are likely to use the pack desire; 'realism' or 'flexibility'

This may sound kind of weird but try lowering the breaking resistances and upping the crash tolerances on the parts that are giving you wobbly issues. I dont know why but for some parts this seems to work very well for fixing the wobble. Its like the high breaking torques values make the part too rigid and instead of flexing & absorbing some if the movement, it all gets transfered through the object till it gets to a week point where it can wiggle, and does so with gusto.

Also, is the wobble occuring only at joints where an engine is located, or is any where you use that part. Reason being that I\'ve found in some cases, the engine it self is the weakpoint and is where the deflection is actually occuring. Adding the breaking resistance lines to engine\'s part.cfg seems to help alot if this is the case. It also seems to help prevent the parts from getting stuck on the launchpad.

[Tiberion]

Its most of the connections - because the additive mass has overwhelmed what the connection engine can handle - all the relevant parts have crashTolerance of 40 and breakingForce/Torque of 100000 - it seems to be a common issue in this 'weight class' of mods.

Backing off to something inbetween those numbers and the earlier one stopped it, and the rockets don\'t get so unruly when so large. Unless everyone is just in love with the idea of the 'stock scale' mass levels I am gonna work on balancing these new settings.

[Thunderbird]

as long as its within a reasonable range of balance with vast majority of the mods, im happy =P

(aka the .3 numbers... wasnt really balanced but close enough)

also,can the instabilities be a result of the collision model with the individual parts themselves?

[SmiteZero]

I\'m fairly sure I heard somewhere that wobble is linked to the collision mesh and the size of the mating faces of adjacent parts.