|
Post by boersgard on Aug 18, 2018 11:00:02 GMT
Well, the current engine so far: 10m x 15m Tons: 47 Thrust: 307 MN Thrust:Mass Ratio: 668 Exhaust Velocity: 9.16km/s Gimbal: 15 Degrees And it accelerates my 5.4kt rocketship at 5.8g, and gives me 9.77km/s delta-v So I've been trying to reduce the acceleration in exchange for better engine efficiency, but I don't actually know how to do that. I assumed it would be as simple as increasing exhaust velocity - but then I'd test out higher ev and my overall delta-v would drop. Then sometimes I'd change things and my TMR might drop 20 units but that would somehow improve my delta-v by 0.20 - 0.50 km/s.
Here was my baseline:
10m x 18m Tons: 67.5 Thrust: 487 MN Thrust:Mass Ratio: 736 Exhaust Velocity: 9.15km/s Gimbal: None
Accelerates my rocketship at 9g, and gives me 9.24km/s delta-v. Which I really don't understand - it has higher TMR and thrust, and virtually equal exhaust velocity, and yet I get less delta-v out of it.
I've attached both to this post in case anyone wants to take a look.
Also, I really wish we had an option for a much smaller gimbal. I don't need 10+ degrees, just 1-3 is enough - especially with this much thrust, yet a 1-3 degree gimbal requires as much mass as a 90 degree gimbal! I'm even forced to maintain 3+ RPM even though 1 RPM will move 3 degrees in about half a second.
More or less what I'm trying to do is maintain >1.5g acceleration, fitting the engine within a 10m radius housing (either with or without gimballing), and maximizing the delta-v as much as possible.
I wouldn't mind doing away with the gimble, but I don't have a lot of room on both sides of my center of mass for lateral RCS thrusters so I'd be stuck with just a set in the nose.
|
|
|
Post by apophys on Aug 18, 2018 16:35:54 GMT
Well, the current engine so far: Tons: 47 Exhaust Velocity: 9.16km/s And it accelerates my 5.4kt rocketship at 5.8g, and gives me 9.77km/s delta-v Here was my baseline: Tons: 67.5 Exhaust Velocity: 9.15km/s Accelerates my rocketship at 9g, and gives me 9.24km/s delta-v. Which I really don't understand - it has higher TMR and thrust, and virtually equal exhaust velocity, and yet I get less delta-v out of it. Adding more dry mass to the ship, with basically the same exhaust velocity, and the same propellant amount, is clearly going to lower the dV you get.
Thrust and TWR are not really related to dV, only to acceleration.
|
|
|
Post by boersgard on Aug 18, 2018 22:46:07 GMT
So I *should* be focusing on the exhaust velocity to maximize delta-v efficiency of the engine?
What's the relation of the injector to the overall design? I noticed that reducing its size seemed to have a direct impact on the general power rating of the engine, with a smaller injector leading to 50 - 150MN engines rather than 300+ MN's as I currently have. Mostly I'm just unclear on what each part of the engine I'm changing is actually going to do to the engine.
|
|
|
Post by apophys on Aug 18, 2018 23:28:04 GMT
So I *should* be focusing on the exhaust velocity to maximize delta-v efficiency of the engine?
What's the relation of the injector to the overall design? I noticed that reducing its size seemed to have a direct impact on the general power rating of the engine, with a smaller injector leading to 50 - 150MN engines rather than 300+ MN's as I currently have. Mostly I'm just unclear on what each part of the engine I'm changing is actually going to do to the engine.
Yes, to make the most out of your propellant, you should be focusing on high exhaust velocity and low mass.
The injector only determines the mass flow rate, i.e. how much propellant you're pumping through the engine each second. The hotter you manage to get that propellant, the higher your exhaust velocity will be (so don't pump too much compared to your reactor, or your heat will be spread thin). You can cut mass from the injector by making it larger and slower-spinning while maintaining the same flow; a good target for the spin is a bit under 100 RPM.
To get a 5.4 kt craft accelerating at 1.5 G, you need about 81 MN of thrust. Higher thrust engines will always be heavier. You might want to take another look at my standard 100 MN HD NTR. Or take a bundle of my 10 MN ones, maybe stripping their gimbals and attaching them all to one new gimbal.
|
|
|
Post by boersgard on Aug 19, 2018 0:06:19 GMT
Thanks, this is a lot of helpful information
This is my current engine:
As I understand it, I want the nozzle temperature and mach in the nozzle to both be as high as possible to get the exhaust velocity as high as possible?
Does regenerative cooling do anything? (It didn't seem to matter what I did with it) What about this 'gas wall' and 'liquid wall'? They also don't seem to matter.
I think I'm next going to try out a stack of engines on the gimbal so I can extend the nozzle and increase mach without ballooning the overall mass, like you suggested.
|
|
|
Post by boersgard on Aug 19, 2018 1:15:47 GMT
Coming back with some findings/efficiency improvements: Anything that rotates really fast seems to require as low a density as possible, with as high a yield strength as possible, so I did a quick check of what materials qualify: Name - Density:Yield Strength ratio Polyethylene - 35 Calcium - 12.8 Reinforced Carbon Carbon - 2.5 Diamond - 2.2 Vanadium Chromium Steel - 1.44 This should mean that a VCS pump/gimbal will offer the highest possible rotation at the lowest possible weight, and yes when I switched from Diamond to VCS for my gimbal and adjusted the RPM to fit the material's stress limits, I almost doubled my gimballing speed. Which means I can start adding more weight to my rocket nozzle and improve the exhaust velocity a little bit more.
I'm now also looking at what improvement I can get out of the injector being made out of VCS instead of Diamond as well. Reducing the injector size/mass for the same fuel flow should be another area of improvement.
Edit: While the gimbal massively improved its RPM with VCS, getting improvements out of the injector proved a lot harder - it seems that as the pump rotation increases, so does the mass. At low RPM's you can add a few hundred RPM and change the mass of the engine by 1-10 tons. Once you start hitting 800+ RPM, a hundred RPM increase can change the engine's mass by 50 tons or more (for this engine - more than doubling the mass before I reached the stress limits of VCS). I was still able to reduce the overall tonnage of my engine with otherwise equal specs eventually and with extensive tweaking, but only from 48.7t to 48.3t with the switch from a diamond injector to a VCS injector
|
|
|
Post by apophys on Aug 19, 2018 3:24:17 GMT
As I understand it, I want the nozzle temperature and mach in the nozzle to both be as high as possible to get the exhaust velocity as high as possible?
Does regenerative cooling do anything? (It didn't seem to matter what I did with it) What about this 'gas wall' and 'liquid wall'? They also don't seem to matter. You want the chamber temperature as high as possible (the upper limit on it depends on the materials you are using; you want to get close to having errors). You seem to be doing well in this regard. You then want the temperature exiting the nozzle to be as low as you can reasonably get it without spending too much mass on a long nozzle. This gets your propellant to do as much work as it can from its expansion. To reiterate, you want to maximize the temperature difference between the chamber and the end of the nozzle.
Regenerative cooling isn't always useful. But it almost never hurts, at least for diamond bells, so I always have it.
I don't bother with the other listed data much. Haven't figured it out.
The significant flaw I see with your design is the low chamber contraction ratio. I'm not sure exactly why, but you are wasting about 20% of your potential thrust and a bit of exhaust velocity by setting it to 1. A good range for it to be is 3-6; higher is better, but puts more stress on the chamber material, which then needs thickening (mass).
|
|
|
Post by boersgard on Aug 19, 2018 4:12:59 GMT
As I understand it, I want the nozzle temperature and mach in the nozzle to both be as high as possible to get the exhaust velocity as high as possible?
Does regenerative cooling do anything? (It didn't seem to matter what I did with it) What about this 'gas wall' and 'liquid wall'? They also don't seem to matter. You want the chamber temperature as high as possible (the upper limit on it depends on the materials you are using; you want to get close to having errors). You seem to be doing well in this regard. You then want the temperature exiting the nozzle to be as low as you can reasonably get it without spending too much mass on a long nozzle. This gets your propellant to do as much work as it can from its expansion. To reiterate, you want to maximize the temperature difference between the chamber and the end of the nozzle.
Regenerative cooling isn't always useful. But it almost never hurts, at least for diamond bells, so I always have it.
I don't bother with the other listed data much. Haven't figured it out.
The significant flaw I see with your design is the low chamber contraction ratio. I'm not sure exactly why, but you are wasting about 20% of your potential thrust and a bit of exhaust velocity by setting it to 1. A good range for it to be is 3-6; higher is better, but puts more stress on the chamber material, which then needs thickening (mass).
I found I was generally exchanging the contraction size for a smaller throat radius, under the general impression that I want the throat radius as low as possible - so if I lower the contraction ratio, I can also lower the throat radius.
I did a quick test and went back to a contraction ratio of 3, and while I increased thrust by 17% I also increased the overall mass by 32% - TMR dropped from 649g to 536g and exhaust velocity increased by only 0.1km/s. I'll keep testing it out to see if I can do better, but it does seem like the lower contraction ratio was giving me better returns despite the lower thrust.
Edit:
If I increase the Thrust:Mass ratio, but maintain the same exhaust velocity, then I should get improved delta-v in spite of the engine massing more? Or only improved acceleration?
I've now got a TMR of 656. but the engine mass is up to 50.9t from 48.3t and the change was small enough I didn't notice any difference in delta-v so I'm not sure if it increased, decreased, or did nothing to dv.
|
|
|
Post by AdmiralObvious on Aug 19, 2018 7:15:39 GMT
Last I looked it up, I'm fairly certain that regenerative cooling feeds some of the fuel through the engine bell, before finally feeding it through the injector. This helps to keep the engine bell cool, and prevent melting of the bell. You don't need to feed very much fuel through the bell to have a noticeable effect. If you overdo the regeneration, you might actually crack the bell due to the difference of temperature of the cryogenic fuel being passed basically adjacent to a flame on the other side of the pipe.
It's just generally a good idea to use it, since it won't up your mass cost, nor price, like it should be doing.
As for the bell itself, you do want the temperature flow chart to look like the way you currently have it if possible. Basically, if the temperature is in the red zone, you're effectively using the fuel expansion. Once it starts getting into the blue range, your fuel has essentially burnt out of it's effectiveness as at that point it's just kind of accumulating at the end of the bell before going into space. Making the bell too much longer will usually result in less efficiency for the mass, but you still might barely edge out a bit of velocity.
Increasing the thrust/mass ratio should deliver greater acceleration. It won't necessarily increase the DeltaV however, since you're pushing more mass.
|
|
|
Post by L5Resident on Aug 29, 2018 13:34:07 GMT
|
|
ghgh
Full Member
Still trying to make kinetics work.
Posts: 136
|
Post by ghgh on Aug 30, 2018 17:33:17 GMT
Well, the current engine so far: 10m x 15m Tons: 47 Thrust: 307 MN Thrust:Mass Ratio: 668 Exhaust Velocity: 9.16km/s Gimbal: 15 Degrees And it accelerates my 5.4kt rocketship at 5.8g, and gives me 9.77km/s delta-v So I've been trying to reduce the acceleration in exchange for better engine efficiency, but I don't actually know how to do that. I assumed it would be as simple as increasing exhaust velocity - but then I'd test out higher ev and my overall delta-v would drop. Then sometimes I'd change things and my TMR might drop 20 units but that would somehow improve my delta-v by 0.20 - 0.50 km/s.
Here was my baseline:
10m x 18m Tons: 67.5 Thrust: 487 MN Thrust:Mass Ratio: 736 Exhaust Velocity: 9.15km/s Gimbal: None
Accelerates my rocketship at 9g, and gives me 9.24km/s delta-v. Which I really don't understand - it has higher TMR and thrust, and virtually equal exhaust velocity, and yet I get less delta-v out of it.
I've attached both to this post in case anyone wants to take a look.
Also, I really wish we had an option for a much smaller gimbal. I don't need 10+ degrees, just 1-3 is enough - especially with this much thrust, yet a 1-3 degree gimbal requires as much mass as a 90 degree gimbal! I'm even forced to maintain 3+ RPM even though 1 RPM will move 3 degrees in about half a second.
More or less what I'm trying to do is maintain >1.5g acceleration, fitting the engine within a 10m radius housing (either with or without gimballing), and maximizing the delta-v as much as possible.
I wouldn't mind doing away with the gimble, but I don't have a lot of room on both sides of my center of mass for lateral RCS thrusters so I'd be stuck with just a set in the nose.
5.4 G is a great way to turn your crew into a fine red paste not to mention 9 G. If you add more propellant tanks, they will increase your DV and lower your acceleration. You only really need less than 1 G of thrust to maneuver. At a typical laser engagement range of 1000km cross section matters very little. Lasers are going to hit your ship with 100% accuracy, just not your hardpoints. No reason not to add more propellant.
|
|
|
Post by cipherpunks on Jan 23, 2019 16:55:56 GMT
Any thoughts on possible feasibility of using solid NTR propellant (960K T melt)? The material in question isn't that hard (27.58 MPa tensile str., 75.153 GPa Young's, 31.716 GPa Shear, 3.5 Mohs). How should this task be approached from storage/feeding PoV - would we store it pre-cut, in pellets, in fine dust form?... I'm talking about Lithium Hydride.
|
|
|
Post by apophys on Jan 23, 2019 19:36:53 GMT
My instinct is to keep it liquid @ 1000 K (with proper thermal insulation) rather than deal with solids. Mechanically pumping liquid over 1600 K is current tech, with ceramics.
Consider using it as a phase-change heat sink for cooling weaponry, because of its very high heat of fusion: ~ 2.85 MJ/kg.
Consider also that LiH & LiH2 are top-tier neutron radiation shielding materials IRL.
Finally, notice that F 2 + LiH is the best-performing chemical rocket in the modded game.
|
|
|
Post by newageofpower on Jan 23, 2019 20:42:32 GMT
My instinct is to keep it liquid @ 1000 K (with proper thermal insulation) rather than deal with solids. Mechanically pumping liquid over 1600 K is current tech, with ceramics.
Consider using it as a phase-change heat sink for cooling weaponry, because of its very high heat of fusion: ~ 2.85 MJ/kg.
Consider also that LiH & LiH2 are top-tier neutron radiation shielding materials IRL. Finally, notice that F 2 + LiH is the best-performing chemical rocket in the modded game. Well, you might get some small performance advantages out of keeping LiH molten, which should be trivial with reactor waste heat, but is complicated by the need to cryocool the Flourine, especially in small packages (i.e. terminal missile stages). Additionally, some uses (cryocooled stealth mine) are contraindicated by keeping the stuff molten; it's worth developing a F/LiH hybrid rocket. IMHO. Also, Oxygen-Beryllium has better theoretical performance, but Beryllium is pretty rare...
|
|
|
Post by cipherpunks on Jan 23, 2019 21:02:17 GMT
it's worth developing a F/LiH hybrid rocket. IMHO. A couple of years ago I modded just that in KSP (with ModuleManager on ProceduralPart), and used it a LOT. Had some relevant papers stashed somewhere which described ongoing R&D into this stuff, but now I couldn't find it on 1st try. Will revisit.
|
|