Engines - old and ARM.

[Kialar]

Hi there.

I've been reading a lot about overpowered ARM engines these days.

A lot has been written about TWRs and thrust in general. But honestly, I didn't feel like the new engines beeing not balanced.

So I figured around with the numbers and wrote down some ideas about it.

The basic question for was: What exactly is a good (comparable) definition of the strenght of an engine.

I assume, most of us would agree, that TWR and ISP are quite good indications.

From my point of view, those two numbers are not really usefull for a general comparison of the engine's efficiency.

Point is, the higher the TWR, the better and the higher the ISP, the better - from a gamer's perspective, I am not a rocket engineer, thus I always wear the glasses of a gamer.

So I thought, the product TWR*ISP gives a quite reasonable number to compare engines.

As those numbers are not very readable, I did a normation on the product TWR*ISP of the Mainsail, what I defined as 1.

From there I did a quick and dirty list of the liquid fuel thrusters.

Putting this into a graph, one will get the following, let's call it efficiency indicator:

After thinking about it, I would more easily say: the new thrusters are more an evolution.

What is your opinion?

[Javster]

There is a slight increase in increase for the last few.

[Daid]

ISP and Weight is what you care about once you are out of the atmosphere. As those new engines are really heavy, I think they are less useful then people would assume. As you would be much better off with the same weight in LV-N engines.

And I don't think they are too strong, your graph shows a nice dip near the 24-77 and LV-T30, I think the engines there are just under-powered compared to the rest, as the rest follows a pretty good curve.

[Starwhip]

Note that this is a Saturn V (Yes, a real one.) and it is on Earth.

From an article:

"The S-IC was built by The Boeing Company at the Michoud Assembly Facility, New Orleans, where Space Shuttle External Tanks would later be built by Lockheed Martin. Most of its mass of over two thousand metric tonnes at launch was propellant, in this case RP-1 rocket fuel and liquid oxygen oxidizer with a fuel efficiency of just under 5 inches per US gallon (just under 4 cm per liter) overall.[15] It was 42 meters (138 ft) tall and 10 meters (33 ft) in diameter, and provided over 34 meganewtons (7,600,000 lbf) of thrust to get the rocket through the first 67 kilometers (220,000 ft) "

A mainsail provides 1.5 kilonewtons

Talk about overpowered...

EDIT:

This is only the first stage, by the way.

[Ravenchant]

A mainsail provides 1.5 kilonewtons

Er. It's 1.5 meganewtons. But still not that much.

[NecroBones]

Er. It's 1.5 meganewtons. But still not that much.

Right, a mainsail is about 1.5 meganewtons, and each of the five F-1 engines on the Saturn-V were about 6.77 meganewtons, with about 260 ISP.

[KerbMav]

Could we stop discussing/comparing KSP to real life rockets? It does not make sense and leads nowhere.

KSP parts have to fit into KSP - fullstop.

It is the same as arguing the size and gravity of KSP planets by citing numbers from our solar system.

Please.

[NecroBones]

Could we stop discussing/comparing KSP to real life rockets? It does not make sense and leads nowhere.

KSP parts have to fit into KSP - fullstop.

Fair enough.

ISP and Weight is what you care about once you are out of the atmosphere. As those new engines are really heavy, I think they are less useful then people would assume. As you would be much better off with the same weight in LV-N engines.

And I don't think they are too strong, your graph shows a nice dip near the 24-77 and LV-T30, I think the engines there are just under-powered compared to the rest, as the rest follows a pretty good curve.

This is an excellent point (emphasis mine). Combining TWR/ISP in the last column of the chart will be a little misleading, since the TWR of the engine by itself isn't that meaningful, as compared to the TWR of the entire spacecraft. Raw weight of the engine matters of course, and you will lose some overall dV efficiency with heavy engines (for instance, one LV-N on a ship will be more efficient than putting four of them on, with the same fuel supply and all other things being equal, since the ISP will be the same, but you will have more weight to push).

Ultimately, the new engines are really only useful for lifting heavy things into orbit, and they excel at that, but they do also face the disadvantage of a worse fuel weight ratio on the larger tanks.

[Z-Man]

So I thought, the product TWR*ISP gives a quite reasonable number to compare engines.

No, it is not. For lifter engines for launches from kerbin, use I_sp*(1-2/TWR).

Rationale: Assume you build a huge asparagus. You already have the N upper stages, total mass M, capable of lifting themselves once state N+1 detaches. Let's design stage N+1 (using negative numbering).

It is well know that for the ascend phase, a total TWR of 2 is optimal. The upper stages have that, so all we need to do is make stage N+1 have total TWR of 2 and we're set.

Assume the engine has mass m and fuel tanks have zero dry weight (for simplicity). We put one engine on the stage (add symmetry later), how much fuel can we take? Since the engine has a TWR of, umm, TWR, and the target TWR is 2, it can lift a total weight of m_t = TWR*m/2, the fuel to take will be m_t-m. How much dv do we get out of that stage, assuming all other engines are also the same or have the same ISP? The rocket equation tells us it's

dv = I_sp * ln((M+m_t)/(M+m)) = I_sp * ln(1 + (m_t-m)/(M+m))

If we assume M >> m_t (a large asparagus), that is approximated via ln(1+x) ≈ x by

dv = I_sp * ln((M+m_t)/(M+m)) = I_sp * (m_t-m)/M

That's what we get out of a stage with that engine. What's the price? Clearly, a good metric is the logarithmic mass increase:

price = ln((M+m_t)/M) ≈ m_t/M

The metric for the engine to use would be dv/price:

Metric = dv/price = I_sp * (m_t-m)/m_t = I_sp * (TWR*m/2 - m)/(TWR*m/2) = I_sp * (1-2/TWR).

Voila.

Of course, you can make other assumptions. These are merely the ones that lead to a concise result that includes both I_sp and TWR and nothing else.

[Vanamonde]

Guys, there are already at least 10 threads on ARM engine balance.

http://forum.kerbalspaceprogram.com/threads/74439-The-Grand-0-23-5-Discussion-Thread

http://forum.kerbalspaceprogram.com/threads/75696-Theory-vs-Practice-The-ARM-parts

http://forum.kerbalspaceprogram.com/threads/75013-I-m-sorry-I-I-just-don-t-like-ARM-%28

http://forum.kerbalspaceprogram.com/threads/75683-ARM-Parts-Poll

http://forum.kerbalspaceprogram.com/threads/75661-The-7-Balance-Points-of-engines-in-a-Full-Career-Mode

http://forum.kerbalspaceprogram.com/threads/75332-Can-both-Career-Mode-and-Sandbox-Mode-be-balanced-simultaneously

http://forum.kerbalspaceprogram.com/threads/75094-Would-you-like-reducing-the-thrust-of-the-new-engines

http://forum.kerbalspaceprogram.com/threads/75738-ARM-SLS-Parts-and-the-Start-of-The-Career-mode-age

http://forum.kerbalspaceprogram.com/threads/75090-the-ion-engine-is-way-too-OP

http://forum.kerbalspaceprogram.com/threads/74603-Engine-balancing-issues-in-ARM

The same people are having the same arguments over and over again, creating a lot of repetition for other members to have to sift through to find information. Please add your comments to one of the existing discussions rather than keep on starting new threads on the same subject.