Solid Rockets Balance Discussion

[Torquemadus]

We know from the latest Devnotes that Ted will be undertaking the daunting task of rebalancing the game, including engine stats.

I wanted to open up a discussion on solid rockets. Specifically, I want to discuss the role they play in heavy lifting.

IRL, in the case of the NASA Space Shuttle and the SLS, the solid boosters are supposed to provide the majority of the thrust at liftoff. Currently, the largest SRB we have available in the stock game is the S1 SRB-KD25k, which produces the equivalent thrust of a Rockomax Skipper. These boosters came with the SLS parts introduced for the Asteroid Redirect Mission pack and are supposed to represent the five segment boosters used on the Block One SLS. The Block Two SLS is supposed to use "advanced boosters", for which solid and liquid fuelled versions have been proposed. The liquid version is represented in game by the LFB KR-1x2. The solid version is not currently represented in stock, I dug up an article about them at nasaspaceflight.com.

In career mode, the early game tech tree is dominated by excellent solid rockets. At the beginning of the game, the RT-10 allows cheap atmospheric and sub-orbital flight. The Rockomax BACC provides cheap first stages for orbital flights and early trips to Kerbin's moons. The S1 SRB-KD25k provides excellent and cheap first stage thrust without even requiring an upgrade to the Research and Development building to unlock it.

All of the current SRBs are 1.25m parts. Currently, there are no larger solid rockets for use in heavy lifting. There are 2.5 metre and 3.75 metre liquid fuelled parts designed for heavy lifting such as the Rockomax Mainsail, LFB_KR-1x2, Kerbodyne KR-2L and S3 KS-25x4. However, there are no solids that produce equivalent thrust. This means that any rockets built to lift large payloads must either cluster large numbers of SRBs, or rely entirely on expensive liquid fuelled first stages.

One issue that will intrude into this discussion will be the game balance of expendable rockets versus SSTOs in the late game. There will also be a question of whether expendable jet lifter stages are competitive against solids.

I personally think that the stock game would benefit from a 2.5 metre solid rocket equivalent to the Mainsail and LFB and available in the Heavier Rocketry tech tree node. Like other solids, it should be cheap, and should have obvious performance disadvantages that make it undesirable for use as anything other than a first stage. Therefore, I think it should have similar thrust to the Mainsail and LFB, should have much worse dry mass and ISP, and should be offered as a cheap "trash bins full of boom" option for a disposable first stage.

[FishInferno]

I also think that the escape tower needs a rebalanced. It is currently underpowered compared to real escape towers, which burn for much longer than the one we have in KSP.

[ghpstage]

In career mode, the early game tech tree is dominated by excellent solid rockets. At the beginning of the game, the RT-10 allows cheap atmospheric and sub-orbital flight. The Rockomax BACC provides cheap first stages for orbital flights and early trips to Kerbin's moons.

The RT-10 is excellent for orbital flights too, it can bring launch costs down to suprisingly low levels. I haven't used them for manned Mun missions since long before the patch however, the parts count and weight limit hurts them.

However, the BACC is nothing but a major step down in my opinion, having around twice the fuel is fine, but it has less than twice the thrust, more than twice the total mass, made up in part by three times the empty mass, and costs more than twice as much for the pleasure.

Its made worse that the RT-10 benefits considerably from stack staging, the probe below almost reaching escape velocity before the SRBs run out.

You can cut the weight further by using a small fuel tank and LV-1 engine, where the SRBs will put it into a solar orbit crossing either Eve's or Duna's without the LV-1 even firing.

You won't achieve this muc with the biggest SRBs, but at least the S1 SRB-KD25k competes favourably in the simple self weight dV category compared against an RT-10.

S1 SRB-KD25k seems respectable, but I can't remember using it much.

Edited January 8, 2015 by ghpstage

[prophet_01]

I agree that the BACC needs a small buff. But just a small. I consider it underpowered but I still use it a lot. I frequently use a first stage of 5 solid boosters, the center one being a BACC and the others depending on the payload's weight. With the part count introduced I don't use radial decouplers that much anymore, at least in early and mid game. BAAC shine if you use FAR and have problems with the part limit. If I manage to keep them stable during the ascent, I get a rly low cost launcher for small payloads. They lift almost all of my probes and comsats and apart from SSTO's there is no real competition for them in that field.

A 2.5 booster would be awesome btw

[marcushouse]

Yes more stock parts such as larger solids would be great. That way we can make this

http://images.nationalgeographic.com/wpf/media-live/photos/000/012/cache/space-shuttle-atlantis_1223_600x450.jpg

[prophet_01]

Uhhh shuttle O___O

but honestly, we still can't. You need high gimbal ssme's for a shuttle of that size (or/and throtle up during the flight with rly big engines) and IIRC the shuttles boosters aren't comparable to the boosters in ksp. They don't increse thrust that fast and then just stop

edit: the shuttle SRB's also had gimbal

[kujuman]

Uhhh shuttle O___O

Did someone say Shuttle?!

Javascript is disabled. View full album

But seriously, my time with AdvSRBs has demonstrated exactly what's been stated: stock SRBs have minimal late game application unless the player just spams the heck out of them. The SRBs on this shuttle are 2m (not 2.5m). Together, they hold 132,900 kg of SolidPropellant (@223-244s Isp), which is 35% of lift-off mass. Together, they give 3797 kN of thrust at lift-off, 66% of total thrust, and provide about 800 m/s of dV over 99 seconds. Did I mention 1.2 degrees of gimbal?

For SRBs to work well in the late game in KSP I think they need a few things:

1. Stack-able. Fuel should burn evenly throughout the SRB stack so all segments burn out at the same time. Also players aren't stuck with only a few size options because they can build longer boosters.1a. Ceteris paribus, boosters with more segments should burn for higher thrust, as there is more surface area of propellant burning. Wider boosters burn longer at lower thrust than narrow boosters of equivalent fuel mass.

2. Gimbal. I think in general gimbal ranges should be increased, but SRBs that lift shuttles or SLS style need gimbaling to function. Shuttles for obvious reasons, SLS style because they now provide a great deal of roll control.

3. Burn profiles, not just constant thrust. These are easy enough to write automatic solvers for since I doubt Squad would want very editable SRBs in stock, because it'd make things too confusing for new players. The SRBs on the shuttle above use an automatic profile that is set to provide constant acceleration (TWR) for the vehicle (if we assume only the SRBs are burning), and that is a very good general purpose setting. No more massive TWR increases at the end of a stage.

[Renegrade]

I've always thought the stock SRBs were way too low in thrust (and I too would like to see the BACC improved over the RT-10 'trashcan'). I think part of the problem is that when SRBs were first introduced, tweakables didn't exist yet, so having a 1.5 MN BACC would result in it burning 98% of it's fuel against the jello-like stock air trying to reach 1,000m/sec at 200m altitude.

Now that we have tweakables, it would be nice if the SRB thrust levels (and BACC dry mass) be revisited.

[Jouni]

I've always felt that the SRBs have too much thrust for their size. I usually tweak them to get the burn times to 60-120 s, depending on whether I want just a quick boost to the initial TWR, or boosters that actually do some useful work.

Of course, because all stock SRBs are so small, they're only useful for launching small to medium payloads.

[THX1138]

I would really appreciate bigger rocket boosters to spend my science points on! This would have the very welcome benefit of reducing lag due to firing 80 SRBs at the same time in the first stage of heavy launches.

Edited January 8, 2015 by THX1138

[prophet_01]

Did someone say Shuttle?!

-snip-

But seriously, my time with AdvSRBs has demonstrated exactly what's been stated: stock SRBs have minimal late game application unless the player just spams the heck out of them. The SRBs on this shuttle are 2m (not 2.5m). Together, they hold 132,900 kg of SolidPropellant (@223-244s Isp), which is 35% of lift-off mass. Together, they give 3797 kN of thrust at lift-off, 66% of total thrust, and provide about 800 m/s of dV over 99 seconds. Did I mention 1.2 degrees of gimbal?

For SRBs to work well in the late game in KSP I think they need a few things:

1. Stack-able. Fuel should burn evenly throughout the SRB stack so all segments burn out at the same time. Also players aren't stuck with only a few size options because they can build longer boosters.1a. Ceteris paribus, boosters with more segments should burn for higher thrust, as there is more surface area of propellant burning. Wider boosters burn longer at lower thrust than narrow boosters of equivalent fuel mass.

2. Gimbal. I think in general gimbal ranges should be increased, but SRBs that lift shuttles or SLS style need gimbaling to function. Shuttles for obvious reasons, SLS style because they now provide a great deal of roll control.

3. Burn profiles, not just constant thrust. These are easy enough to write automatic solvers for since I doubt Squad would want very editable SRBs in stock, because it'd make things too confusing for new players. The SRBs on the shuttle above use an automatic profile that is set to provide constant acceleration (TWR) for the vehicle (if we assume only the SRBs are burning), and that is a very good general purpose setting. No more massive TWR increases at the end of a stage.

Yeah that would be great... a few thoughts of mine: gimbaling should be a tweakable aswell, it also increases the price of the booster. That way they would still remain remotely competable to ssto's (tradeoff: time to complete the ascent vs budget).

I would keep the burn profile the way it is at the begining and add a tweakable for other option(s) once you unlock the big ones (maybe the same unlock for gimbal). Constant thrust would be rly nice, but I do also like the current burn profile for mysmall lifters (I usually reduce the thrust by 10 - 50%).

the possibility to stack would be sooo freqkin awesome the way you mentioned it... I hope that at least some of this makes it into the game eventually

Edited January 9, 2015 by prophet_01

[vexx32]

Gimballing is a tweakable -- looking at the Tweakables menu for any gimbal-capable engine yields the option to Lock Gimbal.

[Whirligig Girl]

This must into game.

[Renegrade]

Gimballing is a tweakable -- looking at the Tweakables menu for any gimbal-capable engine yields the option to Lock Gimbal.

Uhh, current stock SRBs don't have gimballing at all.

http://i.imgur.com/uzxnjtg.png

This must into game.

O.o ...where is that from?

[meve12]

O.o ...where is that from?

Ditto. Wherever that came from I want it.

[Bill Phil]

I think we need radially mounted 0.625 SRBs that are ~ 1 tonne. Basically 1/4 of an RT-10 but only radially...

[kujuman]

Gimballing is a tweakable -- looking at the Tweakables menu for any gimbal-capable engine yields the option to Lock Gimbal.

I think that's more of the idea that gimballing is an after market addition to an engine nozzle, and its inclusion on any particular engine is tweakable (for a cost/mass penalty).

[THX1138]

Do real-life SRBs have gimball or post-ignition adjustable thrust?

[vexx32]

I'm not sure I've heard of any gimballed SRBs, but it's by no means impossible to do.

As for adjustable thrust, not so much. What Gregrox has there is what's called a, uh... damn, what was it called again. Thrust profile? No, no, something else... ah! Grain geometry! Basically, depending on how the solid fuel in the SRB is arranged, the engine will burn in a different manner.

A solid rocket booster works by having the fuel packed around the inside edge of the container, leaving the centre free to allow air to be present upon ignition, which is necessary to burn the fuel. Typical SRBs will have the hollow core almost all the way through the length of the cylinder, to keep a nice even burn and prevent awkward weight distributions from occurring as fuel is burnt. Having the centre cavity be a different shape allows engineers to design an SRB that might start out with a very high thrust, but quickly peter out as the surface area of the burning fuel is reduced. Essentially, it permits them to modify the thrust to behave in a particular way before the SRB is ever ignited.

While perhaps a slightly technical feature, it would -- I'm sure -- have a lot​ of uses in KSP.

[Rdivine]

I also think that the escape tower needs a rebalanced. It is currently underpowered compared to real escape towers, which burn for much longer than the one we have in KSP.

I attach 8 sepatrons on the top of the tower, with a thrust limit of 50.

4 is used for pulling away the crew module,

the remaining 4 is fired when the tower is jettisoned.

You can make the design look neat with the new widgets.

[Torquemadus]

I've finally got to the point in my 0.90 hardcore custom difficulty save where I have enough tech to use the lightest version of my Kerbooster.

Here are some pics of launch vehicles used in my current hardcore save:-

This is Kercury 8, which is a good example of my BACC booster based early game launchers.

This is Kercury 11. I used SRBs to send a number of these ships to spam contracts from both of Kerbin's moons.

This is the Kerbooster Light, which uses SRBs to perform low cost satellite launches.

I can launch larger versions of the Kerbooster, but I have to spam large numbers of SRBs as a first stage.

The current SRB-KD25k is too weak for heavy lifting. It can only be used to lift large payloads if clustered in large numbers. A larger SRB is needed for use with the 2.5m and 3.75m parts.

[prophet_01]

The shuttle's SRB's had gimbal ( http://en.m.wikipedia.org/wiki/Space_Shuttle_Solid_Rocket_Booster ).

I was thinking of a gimbal tweakable that is tied to a cost and mass penalty, so gimbal is an option but comes at a price. I think that cheap boosters without gimbal should still be part of the game, but sometimes I would like to trade cost- and fuel efficency for conrol

The thrust depends on the surface area of the solid fuel within the booster. Depending on the shape of the fuel within the casing the thrust may vary during the burn. There are pictures of some examples on wikipedia

That mod looks rly useful. It would be nice to have this in stock. I'd prefer to unlock it with a late-game tech tree node. No need to confuse new players

[Renegrade]

I think we need radially mounted 0.625 SRBs that are ~ 1 tonne. Basically 1/4 of an RT-10 but only radially...

Very much so, please! (Plus some 2.5m ones..)

Do real-life SRBs have gimball or post-ignition adjustable thrust?

Yeah, see below (SRBs are actually used extensively in rocketry these days due to cost factors) for gimbals. Also, Vexx32 covered static thrust profiles (those would only be changeable in the VAB like the thrust tweakable for SRBs), but there's also hybrid SRBs (I believe that SpaceShipTwo uses a hybrid SRB design), wherein the fuel is solid like a conventional SRB, but the oxidizer is a liquid, and thus the engine can be turned on, off, or throttled as needed.

I don't think we need hybrid SRBs, but I'd absolutely love to be able to set a profile for the existing SRBs

The shuttle's SRB's had gimbal ( http://en.m.wikipedia.org/wiki/Space_Shuttle_Solid_Rocket_Booster ).

Not just gimbal - LOTS of gimbal. It's like eight degrees. I don't think there's any stock engine with eight degrees of gimbal.

Also the SRBs provide the vast majority of the thrust - something like 26MN total vs. only 5-ish for the liquid fuel engines - for the Shuttle, and are far, far more powerful than ANY liquid fuel engine ever made (13-14mn ea)

This is the Kerbooster Light, which uses SRBs to perform low cost satellite launches.

http://i.imgur.com/QxCMhPz.jpg

http://i.imgur.com/krRFM5P.jpg

http://i.imgur.com/K3xa1HA.jpg

http://i.imgur.com/FobnDZj.jpg

The Kerbooster Light is very much like my own stock light lifter design.. and the Kercury 8 similar to some BTSM designs I use (only I've never used above 4-way symmetry for those).