How to play without asparagus staging?

[Hysterrics]

38 minutes ago, cantab said:

This is a common approach for me as well. Simple, like a real rocket, and burning the core engine from the start (sometimes throttled back) gives me crucial thrust-vectoring for control, since KSP doesn't have vectored SRBs.

Well I'd say it's not quite so cut-and-dried. Most real rockets feature two payload ratings - one for LEO, and a second lower rating for Geostationary Transfer Orbit. Sometimes other configurations are used too, like Saturn V where the third stage put the spacecraft on Trans-Lunar-Injection. Or New Horizons, where the Atlas V's second stage reignited in low orbit to start the ejection burn (that was completed by a Star 48 solid motor which you can argue was part of the payload).

What I think is more generally the case is the launcher is doing its burns around or below low orbit. Sometimes there's a wait and reignition in low orbit, but once you start thinking about the burn to enter geostationary orbit or the burn to capture at your destination planet or moon, that's definitely something that's the job of the payload. Though since KSP lacks fuel boiloff and part reliability concerns, if I have excess delta-V in my launcher I'm probably gonna keep it with me anyway!

I suppose it just depends on you want to do. But often time the first stage gets you through most of the atmosphere. Which is what I was trying to say.

[Eric S]

1 hour ago, 5thHorseman said:

No, it's as true as I think. If you ever get the chance alone on a merry-go-round, get on the edge and slowly make your way around it. You'll notice that the Merry-go-round more slowly (but perceptibly) goes the other way.

Even though there is no outside force on the two of you, your force on it (or more easily imagined the shifting of the center of mass of you two as a pair) causes - when one of you moves relative to the COM - for the other to move to compensate.

Does the merry-go-round continue to accelerate as you walk around it?  No.  The only force was applied when you accelerated.  Now, stop.  The merry-go-round stops because your deceleration applied a force opposing the original force.  Now, picture the fuel line shifting fuel.  At any given time after the fuel line is primed, there's a balance between the force applied by the fuel accelerating into the fuel line and the force applied by the fuel decelerating as it comes out of the fuel line.  Yes, there's a kick as the fuel lines fill, but the launch clamps could absorb that kick.

 

 

[GoSlash27]

The problem of whether asparagus staging would actually induce a roll IRL is fascinating, but off- topic.

 I think it's like an airplane taking off from a treadmill.

Maybe we should send it in to MythBusters?

Best,

-Slashy

Edited January 1, 2016 by GoSlash27

[Superfluous J]

1 hour ago, Eric S said:

Does the merry-go-round continue to accelerate as you walk around it?  No.  The only force was applied when you accelerated.  Now, stop.  The merry-go-round stops because your deceleration applied a force opposing the original force.  Now, picture the fuel line shifting fuel.  At any given time after the fuel line is primed, there's a balance between the force applied by the fuel accelerating into the fuel line and the force applied by the fuel decelerating as it comes out of the fuel line.  Yes, there's a kick as the fuel lines fill, but the launch clamps could absorb that kick.

Well the merry-go-round isn't a perfect analogy because you're not, at the end of your journey, being ignited and spat downward at the greatest possible velocity. I can't think of a real-world way to simulate that, other than building something that holds water and allows it to flow out as if the thing were an asparagus setup for a rocket. I'm not about to build such a contraption

If the fuel was just getting pumped in a circle, sure, only torque would be applied at the beginning, and then torque backward would be applied at the end of pumping. But the fuel is being pumped around in a half circle and then pumped inward, where it then goes straight down.

[sardia]

Did you know that a single decoupler can detach 8 SRB boosters?  You can't beat that for cost. All my first stages are powered be rcs, srb and SAS alone. Plu work well if you have a areodynamic ship. 

[Sovek]

my typical rockets use a center stage and side boosters really.

 

A 1.35m rocket (Procedural Fuel Tanks) with clustered 45kn engines (x4) with over 8km/s of dV, 3 stages, probe itself has over 2.5km/s. Used for putting com sats into a 5.5Mm x 5.5Mm 0 degree orbit in 3x harder solar system.

 

When I was playing with KScale2 (2x, no inclination changes) and 8km/s would get me to the Mun, and back, I took a page out of the R7 book. 6 Wildcat V (KW engines) powered this thing on ascent. The core had more fuel but all 6 burned about evenly which left 1km/s of dV in the tank. In stock you could just about get into orbit without the second stage.

 

SRBs will sometiems be used, it depends. Generally I imitate what works IRL and it usually can work in KSP, quite well.

Main tips.

1. use efficient engines, the most efficient you can find.

2. Weight is important. That commsat was mostly fuel, aside from a little bit of hydrazine, 8 RCS ports, a probe core, battery, solar panels and an antenna. The weight of the battery killed more of my dV than anything else, combined. on long range probes, weight is everything.

3. Use side boosters (just without the fuel ducts) and tweak engines thrust to match what you need. Don't be afraid to take a page out of the Saturn 5 and do a center engine cutoff to keep the T:W in check. Burning in one long go is more efficient than burning until you get a Ap of 70km +. The only time I'm  "cruising" nowdays is staging.

4. In most cases, engine clusters are your friend. I use real fuels so your experience may vary, but 1.25m engines usually beats one large 2.5m engine. This is especially true if 2 1.25m engines provide the correct thrust I need than that 2.5m engine that I have to throttle way down so as to not fly off the pad at mach 1.

5. Limit your T:W, 1.15-1.2 at the pad is plenty for most cases. Faster can just result in all sorts of problems, either fighting drag or causing rockets to flip because your moving too fast, too soon.

[Xavven]

11 hours ago, 5thHorseman said:

I f the fuel was just getting pumped in a circle, sure, only torque would be applied at the beginning, and then torque backward would be applied at the end of pumping. But the fuel is being pumped around in a half circle and then pumped inward, where it then goes straight down.

Hm I think Eric has a point.

If the velocity vector of the fuel is this:

----->

What force must you apply to end up with a velocity vector of this (to pump it down into the engine where it is then burned)?

|

v

 

I think the answer is you accelerate the fuel to the left and down. This is equivalent to stopping on the merry-go-round.

[Eric S]

Xavven, yes, that's the point.  The equal and opposite reaction of making the fuel stop moving to the right and move down would cause the merry-go-round to stop spinning (or in this case, counteract the force of the fuel that is starting to go around at that particular moment) and accelerate up.

As was pointed out, this is off topic, so I'm in favor of moving the discussion to private messages.

[Geschosskopf]

On ‎12‎/‎31‎/‎2015 at 9:19 PM, 5thHorseman said:

THAT is what is not super realistic for at LEAST the reason that it would introduce torque in the ship, pumping fuel around in a circle like that.

My problem with asparagus, and even crossfeed, is that KSP's fuel flow system makes them work way better than they would in real life.  If you take what KSP fuel lines do at face value, that leads to rather silly implications.  For example, there's no limit on the flow rate of KSP fuel lines.  No matter how fast the downstream engine, be it a Spark or a Mammoth, is burning fuel, the line from the upstream tank always delivers fuel at exactly the same rate, so the downstream tank stays completely full until the upstream tank is empty.  That's asking a lot of a small pipe of constant diameter which might be rather long in some extreme situations. :).

But it's asking even more of the fuel pumps.  The fuel line pump has to deliver whatever flow rate the downstream engine needs against a considerable and constantly increasing backpressure caused by the downstream tank remaining full while the upstream tank drains at twice the natural rate.  And it has to keep doing this even down to the last drop of fuel in the upstream tank, all while not interfering with the flow going to the upstream tank's own engine.  That's an amazing pump there, especially because the same small, light unit works for whatever size engines are involved

I muse on such things because I'm a firefighter so I pump a lot of fluid through small, flexible, yellow tubes of arbitrary length .  This involves dealing with math about water pressure and flow rates so I have some passing familiarity with the subject.  Anyway, I mostly deal with pumps that can move up to 1500 gallons per minute.  With water being about 8.4 pounds / gallon, tha's 12600 pounds per minute being moved.  A pump that can do this is rather large.  It spans the entire width of the big red truck required to move it around and needs several hundred diesel horsepower to spin it.

Then I look at the stats of the F-1 engine's fuel pump:  15,471 gpm of RP-1 and 24811 gpm of LOX for a total weight moved of 340980 pounds per minute.  Even scaled down to KSP size, that's still 3662 combined gpm, and 30998 pounds per minute, both more than twice the capacity of my $400K, 20-ton firetruck.  And that's just 1 engine---the Saturn V had to feed 5 of these at once.  Then if you put such boosters in a KSP asparagus arrangement to lift some monster payload, each tank in the chain would need twice this pumping capacity, one for itself and the other for the fuel line.  Egad!

Of course, the F-1 fuel pump was small and light, and used a gas turbine with 55000 bhp, so was a totally different beast than a fire pump.  However, I still find it illuminating to think of each KSP-sized F-1 analog as needing 2.5 full-size firetrucks just to feed it, and 7.5 firetrucks for every pair in an asparagus arrangement (2.5 in the dlownstream tank, 5 in the upstream tank).  This mental image is enough for me to call BS on KSP fuel lines.

[Scarecrow]

My Apollo emulator doesn't use asparagus staging, and I don't think it's too tall as a result of not using it.

 

[GoSlash27]

29 minutes ago, Scarecrow said:

My Apollo emulator doesn't use asparagus staging, and I don't think it's too tall as a result of not using it.

 

Ultimately, asparagus staging is overrated in KSP anyway. At least for the purpose of cheaply putting payloads into LKO.

 a simple 2 stage lifter can do the job cheaper and is much easier to design.

The Asparagus has too many staging events for the DV need and uses clusters of engines that wind up costing more in total than the single engine they replace. Plus the cost of decouplers, fuel lines, nose cones, possibly struts and sepratrons...

Fiscally it's just not worth it. I suppose that's why nobody uses them anymore...

Best,

-Slashy

Edited January 1, 2016 by GoSlash27

[Temstar]

16 minutes ago, GoSlash27 said:

a simple 2 stage lifter can do the job cheaper and is much easier to design.

Don't be so sure about the cheaper part. I'm getting numbers approaching √800 per ton to orbit using a recoverable upper stage in my Nova II. The fact that asparagus staging is efficient, particularly it's efficient in number of engines a rocket need works to drive the cost down by a lot.
 

These days asparagus staging rockets are much easier to design than the bad old days. One of the key problem with asparagus staging rockets back in the days was lack of roll control, and modern KSP has made this issue trivial by:

1. Stronger joints, so booster around the core flex much less under thrust. Previously you need to be know a lot about the art of strutting to make the whole structure stiff, now just a few struts radial coming out from the core to the boosters will do
2. Vector thrust can now be used for roll authority, Previously to control roll you could only rely on reaction wheel and RCS. Control surfaces don't work for roll when you needed them the most (after gravity turn) and roll authority vectored thrust wasn't invented yet.

 

Edited January 1, 2016 by Temstar

[Zoidos]

14 hours ago, Eric S said:

Does the merry-go-round continue to accelerate as you walk around it?  No.  The only force was applied when you accelerated.  Now, stop.  The merry-go-round stops because your deceleration applied a force opposing the original force.  Now, picture the fuel line shifting fuel.  At any given time after the fuel line is primed, there's a balance between the force applied by the fuel accelerating into the fuel line and the force applied by the fuel decelerating as it comes out of the fuel line.  Yes, there's a kick as the fuel lines fill, but the launch clamps could absorb that kick.

 

 

In principle, 5th Horseman is right. If you stand on a merry-go-round and pump water in a circle into the same bucket where it came from, you are still moving mass around, thus exerting force. This will cause the merry go round to accelerate as long as water is pumped. Although compared to a 100t spacecraft the force of the fuel being pumped around is negligible.

[GoSlash27]

17 minutes ago, Temstar said:

Don't be so sure about the cheaper part. I'm getting numbers approaching √800 per ton to orbit using a recoverable upper stage in my Nova II. The fact that asparagus staging is efficient, particularly it's efficient in number of engines a rocket need works to drive the cost down by a lot.
 

These days asparagus staging rockets are much easier to design than the bad old days. One of the key problem with asparagus staging rockets back in the days was lack of roll control, and modern KSP has made this issue trivial by:

1. Stronger joints, so booster around the core flex much less under thrust. Previously you need to be know a lot about the art of strutting to make the whole structure stiff, now just a few struts radial coming out from the core to the boosters will do
2. Vector thrust can now be used for roll authority, Previously to control roll you could only rely on reaction wheel and RCS. Control surfaces don't work for roll when you needed them the most (after gravity turn) and roll authority vectored thrust wasn't invented yet.

 

Temstar,

 I'm sure a lot of the savings you're seeing are due to economy of scale. Asparagus staging becomes cost effective when you have no choice but to use clustered engines.

In the case where single boosters are sufficient, My "cheep 38" can put 38 tonnes into orbit without bothering to recover anything for $842/ tonne. With upper stage recovery, the price would drop to $557/tonne.  Asparagus can't compete with that in that scale.

Best,

-Slashy

 

 

 

[Eric S]

18 hours ago, 5thHorseman said:

1 hour ago, Zoidos said:

In principle, 5th Horseman is right. If you stand on a merry-go-round and pump water in a circle into the same bucket where it came from, you are still moving mass around, thus exerting force. This will cause the merry go round to accelerate as long as water is pumped. Although compared to a 100t spacecraft the force of the fuel being pumped around is negligible.

 

 

I have no clue why the editor is nesting that.  Sorry 5th, especially if the forum notifies you that I'm quoting you.

I've got a lot of respect for 5th Horseman, but he isn't right in this case.  Something in motion doesn't apply a force.  Accelerating something so that it does move does.  However, so does decelerating.  As long as the amount of fuel moving into the fuel line is equal to the amount of fuel leaving the fuel line and the kinetic energy of the fuel leaving the fuel line is absorbed back into the craft, they balance out.  If the circular motion of the fuel is redirected downward, then THAT force counterbalances the force generated by fuel accelerating into the fuel line.

If pumping water out of a bucket and through a loop back into the same bucket caused rotational acceleration, then why do we use reaction wheels for attitude control, given that those have saturation issues?

On topic, I find I stopped using asparagus staging as much when we got the ARM parts, and have almost completely stopped using it in 1.0.  Bigger engines made asparagus less appealing.  Needing less delta-v to get to orbit, even more so.  I still do liquid boosters feeding a central stack fairly often, but almost never more than one pair of boosters.

 

Edited January 1, 2016 by Eric S

[GoSlash27]

All,

 I made a new thread regarding "asparagus torque" over in the Science forum.

It's a mind- bender and I'm curious about the answer...

Best,

-Slashy

[suicidejunkie]

My career mode heavy rocket policy has evolved into using three stage designs.

First, a set of thumper SRBs for the initial kick up to 250-300m/s, and then drop off.  This is done vertically to avoid damage.  After separation, gravity turn may begin.

Second is three mainsails with 2 popsicles of fuel each.  With payload, these start out <1 TWR, but I don't want to bang against the sound barrier anyways (the payload is typically terrible aerodynamically, but the vehicle remains subsonic until that is not a problem).  As they burn down and atmo thins, speed picks up naturally.

Originally it was intended to put 30 inert tons in orbit, but I usually have at least 60 tons on the top (image is 160t payload), so now the popsicles drop into booster bay for 85% recovery.

 

The payloads themselves do the orbital insertion using their efficient space drives, typically poodles and terrier/909s.

Burning payload fuel to reach orbit isn't a concern for me; they often launch with excess empty tankage anyways, and can either fill up at the orbital depot, or go straight to the Minmus refinery.

Basically, everything that touches space stays there or lands (hopefully) in one piece.

Spoiler

 

[pandaman]

I haven't used asparagus much at all, apart from playing with it to see how it works.  

I found it fun to play with, but a bit fiddly, and as mentioned above, with all the larger ARM parts now it's probably not as necessaryas it was 'way back when'.  

It's a neat concept though, and I did learn some good stuff by trying it.

[GoSlash27]

6 hours ago, Geschosskopf said:

My problem with asparagus, and even crossfeed, is that KSP's fuel flow system makes them work way better than they would in real life.  If you take what KSP fuel lines do at face value, that leads to rather silly implications.  For example, there's no limit on the flow rate of KSP fuel lines.  No matter how fast the downstream engine, be it a Spark or a Mammoth, is burning fuel, the line from the upstream tank always delivers fuel at exactly the same rate, so the downstream tank stays completely full until the upstream tank is empty.  That's asking a lot of a small pipe of constant diameter which might be rather long in some extreme situations. :).

But it's asking even more of the fuel pumps.  The fuel line pump has to deliver whatever flow rate the downstream engine needs against a considerable and constantly increasing backpressure caused by the downstream tank remaining full while the upstream tank drains at twice the natural rate.  And it has to keep doing this even down to the last drop of fuel in the upstream tank, all while not interfering with the flow going to the upstream tank's own engine.  That's an amazing pump there, especially because the same small, light unit works for whatever size engines are involved

I muse on such things because I'm a firefighter so I pump a lot of fluid through small, flexible, yellow tubes of arbitrary length .  This involves dealing with math about water pressure and flow rates so I have some passing familiarity with the subject.  Anyway, I mostly deal with pumps that can move up to 1500 gallons per minute.  With water being about 8.4 pounds / gallon, tha's 12600 pounds per minute being moved.  A pump that can do this is rather large.  It spans the entire width of the big red truck required to move it around and needs several hundred diesel horsepower to spin it.

Then I look at the stats of the F-1 engine's fuel pump:  15,471 gpm of RP-1 and 24811 gpm of LOX for a total weight moved of 340980 pounds per minute.  Even scaled down to KSP size, that's still 3662 combined gpm, and 30998 pounds per minute, both more than twice the capacity of my $400K, 20-ton firetruck.  And that's just 1 engine---the Saturn V had to feed 5 of these at once.  Then if you put such boosters in a KSP asparagus arrangement to lift some monster payload, each tank in the chain would need twice this pumping capacity, one for itself and the other for the fuel line.  Egad!

Of course, the F-1 fuel pump was small and light, and used a gas turbine with 55000 bhp, so was a totally different beast than a fire pump.  However, I still find it illuminating to think of each KSP-sized F-1 analog as needing 2.5 full-size firetrucks just to feed it, and 7.5 firetrucks for every pair in an asparagus arrangement (2.5 in the dlownstream tank, 5 in the upstream tank).  This mental image is enough for me to call BS on KSP fuel lines.

Well... to be fair turbopumps are incredibly light and powerful assemblies, so you can't really compare them to fire engine pumpers.

 

But yeah... I picture all the complex flows and pressures, remember that oxidizer requires a parallel circuit, and all the connections have to positively seal under pressure when preceding stages are jettisoned...

 Getting that to work would be a daunting task. Getting it to work reliably would be a nightmare!

Edited January 1, 2016 by GoSlash27

[Temstar]

6 hours ago, GoSlash27 said:

In the case where single boosters are sufficient, My "cheep 38" can put 38 tonnes into orbit without bothering to recover anything for $842/ tonne. With upper stage recovery, the price would drop to $557/tonne.  Asparagus can't compete with that in that scale.

Can you show me this rocket design?

I want to asparagus seven of them

[Geschosskopf]

8 hours ago, GoSlash27 said:

Fiscally it's just not worth it. I suppose that's why nobody uses them anymore...

I can think of another reason why the Age of Asparagus is over.  Basically, the problem that asparagus was developed to solve no longer exists.  Back in the hayday of asparagus, we only had 2.5m rockets and the BACC SRB, and needed 4500m/s to reach LKO.  Beyond a certain relatively low payload mass, putting that much dV in a 2.5m lifter almost forced you to use asparagus.  And in fact these limits were such that there was much discussion about lifters, challenges to get full orange tanks to orbit, and such.

But now we've got stock 3.75m parts, a slightly bigger (but still too small) stock SRB, and, more importantly,it only takes 3200-3400m/s to reach LKO.  These 2 changes took out both legs that asparagus was standing on, limited rocket size and high dV requirments.  So now it's not hard to build a non-asparagus lifter that can lift much heavier payloads than was really practical back in the day.The fact that this is way less complex and also cheaper than doing asparagus was certainly the last nail in asparagus' coffin, but I daresay that if asparagus was still even remotely close to being necessary today, we'd still be seeing a lot of it.

[Norcalplanner]

6 hours ago, Temstar said:

Can you show me this rocket design?

I want to asparagus seven of them

That might be a bit difficult - the first stage is only SRBs.

[Temstar]

27 minutes ago, Norcalplanner said:

That might be a bit difficult - the first stage is only SRBs.

I still want to see, the numbers seem so good that it's a bit hard to believe. If the numbers are true surely there will be something I can learn from it.
 

[GoSlash27]

13 hours ago, Temstar said:

Can you show me this rocket design?

I want to asparagus seven of them

By all means, but you won't be able to. The first stage is solid propellant

http://wikisend.com/download/380030/cheep38.craft

Best,

-Slashy

[GoSlash27]

Wow, I hadn't seen Honest Jeb's design before. We're both working the same way

 Best,

-Slashy