Feasibility/Interest in an Engine Design Mod?

[Headhunter09]

There are a lot of part packs that provide various sizes and designs of engines. You have vacuum engines, first-stage engines, high-thrust, high-efficiency, dual-nozzle, tall, short, nuclear, fusion, etc.

Since there are mods to introduce extra realism elements to the design process (life support, heat management, communication, safe reentry, aerodynamics, etc), why not create a mod for designing engines?

On the one hand, it would allow the vast amount of available engines to be relegated to a "Preset" list, freeing up the propulsion tab. It would also allow designers to make engines with just enough thrust, and most efficient at the altitude they are likely to be used at.

On the other hand, designing engines is REALLY COMPLEX. How deep do you go? Do you restrict the choices to big things like style and number of turbopumps, cooling cycle design, nozzle vs aerospike, and number of combustion chambers? Do you have a range of values for nozzle angles and sizes; fuel flow rates; valve reliability (tradeoff between failure rate and mass)? Then there's fuel choice, and fuel tank management. Choosing a gas for pressure-fed engines, not to mention arranging the geometry of parts for optimal heat release. Engine design involves fluid dynamics, thermodynamics, mechanical engineering, and material science. Would a mod ever be able to do it justice?

Skimming over the nitty gritty details of how such a mod would be implemented in the current game, is this what people are interested in? There are some people who play KSP, others to design. Is this too real?

Thoughts?

[Javster]

KW has multiple-nozzle engines.

[quietsamurai98]

I've always wanted the R&D part of KSP to actually include Development...

Idea for engines: Unlock types of fuel (Liquid, Air-Breathing, Nuclear, Ion, etc), Thrust modifiers, ISP modifiers, etc. via the tech tree, then mix and match them to develop custom engines.

[Headhunter09]

KW has multiple-nozzle engines.

Most part packs do. What I'm suggesting is the ability to create new engines, deciding how many nozzles, whether each nozzle has its own combustion chamber, how to cool the engine, how to pump the fuel into the combustion chamber, and how big/what size to make the nozzle. In this way you could create a new engine design tailored to your needs.

[WhiteWeasel]

The closest thing I can see so far is the AIES mod that has fuel tanks with multiple attachment nodes allowing for mix and match engines.

[rhoark]

I am working on this. Existing engine parts will be nozzles only and determine only the exit area, gimbal range, and cooling method (regenerative, ablative, radiative). Valves, turbines, pumps, and manifolds will be independent parts. Any part, including tanks, will receive arbitrary resources, and contents will change phase or combust as appropriate. Direction and rate of flow depends on the pressure gradient. Since each tank is a single space, separate tanks will be needed for each resource. Expected use is to surround everything with procedural fairings or 6S bays.

[MKSheppard]

I actually have been working sort of on something like this for the last month -- but it's all basically C++ code that someone would need to convert to C# for use with KSP.

I've been using a simple 1D nozzle modelling routine that was done by Tom Benson of NASA Glenn back in 2005 for a Web Java App -- I converted it to C++ and have been extending it since then.

I recently managed to hack together a crude approximation of modelling the performance characteristics of a annular aerospike (aka plug nozzle); and I'm currently engaged in "modularizing" my code; so that I can reuse each module over different programs.

[MKSheppard]

cooling method (regenerative, ablative, radiative).

Roark; be careful with this; I found out some interesting facts -- engines can use more than one method in the same engine -- for example; the F-1 IIRC uses a regeneratively cooled thrust chamber and regeneratively cools the bell nozzle out to expansion ratio 10:1.

Below that point -- from 10:1 to 16:1; the engine basically used the turbopump turbine exhaust to film cool the nozzle expansion; which resulted in a lot of propellant theoretical energy being lost, making the F-1 quite inefficient compared to say the H-1, which was fully regenerative.

[rhoark]

I won't be modeling different sections of the nozzle separately, but the parts will be modular enough to choose what resource(s) and how much of it to route to cooling, and what to do with the hot/expanded products afterward. The intent is you'll be able to implement whatever engine cycle you can conceive, whether traditional (pressure fed, expander, etc.) or wacky (tripropellant, nuclear preheater, variable mix ratio...).

Edited February 3, 2014 by rhoark