I don't think so. Scaling the dimensions of a rocket engine as not so simple as the dynamics within the combustion and expansion chambers behave differently. IMHO they are more likely to keep the same design and titrate the chamber pressure and thrust upward, like they did with Merlin.
Predicting the optimal size of an engine before the engine is built is the most difficult part.Everything has to be taken into account including what you mentioned above.I think SpaceX got lucky that the optimal engine size came out smaller in real life tests than predicted beforehand.(E.g. Isp was measured higher than predicted for the small Raptor version)
Quote from: Peter.Colin on 09/16/2017 09:27 amPredicting the optimal size of an engine before the engine is built is the most difficult part.Everything has to be taken into account including what you mentioned above.I think SpaceX got lucky that the optimal engine size came out smaller in real life tests than predicted beforehand.(E.g. Isp was measured higher than predicted for the small Raptor version)Link?
Quote from: RotoSequence on 09/16/2017 10:54 amQuote from: Peter.Colin on 09/16/2017 09:27 amPredicting the optimal size of an engine before the engine is built is the most difficult part.Everything has to be taken into account including what you mentioned above.I think SpaceX got lucky that the optimal engine size came out smaller in real life tests than predicted beforehand.(E.g. Isp was measured higher than predicted for the small Raptor version)Link?For instance:http://www.thespaceshow.com/show/22-jun-2017/broadcast-2934-ms.-gwynne-shotwell(@ 40.08)
Do you believe that the thrust to weight and ISP are heavily dependent on the thrust level? I can assure you they are not. The proper size for a rocket has way more to do with mission requirements, flight rates, development and production costs. If I have misunderstood your position, my apologies.John
... Probably a 2000KN Raptor is the same size or even smaller than a Merlin 1D engine.An around 12 meter diameter BFR rocket will have around 75, 2000kN engines......
Quote from: Peter.Colin on 09/17/2017 10:09 am... Probably a 2000KN Raptor is the same size or even smaller than a Merlin 1D engine.An around 12 meter diameter BFR rocket will have around 75, 2000kN engines......Gwynne actually said 'by a factor of 2, up to a factor of 3' times the 1000kN Raptor is optimal.Splitting the difference, let's say 2.5 times is optimal.The 1000kN Raptor is about the same diameter as a Merlin 1D, i.e. 0.89m.The 3050kN Raptor is 1.51m.Thrust is proportional to nozzle area, so for 2500kN, the diameter would be about 1.37m, much larger than Merlin 1D.A 0.75 (9m) scale model of BFR would have 128MN * 0.422 = 54MN thrust.54MN / 2500kN = 21.6 engines, let's round it down to 21.For the 12m BFR it would be 128MN / 2500kN = 51.2, say 48 engines.Both configurations provide excellent packing geometry, and could look something like this:
Engine nos. getting crazy here. Far too many highly stressed parts that could potentially RUD. Also even if no engines fail during a mission, the maintenance costs between flights will be higher with all those engines to check. Anything more than around 20 engines is really pushing it and hopefully EM will keep the engine no. the same from mini-BFR to the future larger versions by dev. larger versions of Raptor. Don't need the absolute best TWR from a booster engine since it's job is to just get the launch stack out of Earth's atm. It's the US engines that need the best possible TWR.
Peter,- Engine T/W does not go up as thrust goes down. It just doesn't. T/W tends to be about constant from 100K lbs to 1M lbs thrust class. ISP isn't much effected by thrust either. - You also seem to be obsessed with filling the base of the vehicle with rocket exhaust. You want to find the optimum vehicle thrust to weight (usually around 1.25). Filling the base is not that important.John
I believe that Space X will develop a larger engine for the full scale, 12M, BFR. FWIW, 48 engines just seems way to many. John
Peter,.......You also seem to be obsessed with filling the base of the vehicle with rocket exhaust.......
Quote from: livingjw on 09/17/2017 07:26 pmPeter,- Engine T/W does not go up as thrust goes down. It just doesn't. T/W tends to be about constant from 100K lbs to 1M lbs thrust class. ISP isn't much effected by thrust either. - You also seem to be obsessed with filling the base of the vehicle with rocket exhaust. You want to find the optimum vehicle thrust to weight (usually around 1.25). Filling the base is not that important.JohnHi John,- For other rockets you're probably right, for the Raptor I think it's different, and to my opinion an advantage that will leave other rocket companies like Blue Origin behind in the dust.- I'm obsessed with filling the base yes, because I know my rocket can be morI e heavy/higher.Rocket weight = combined thrust devided by 1,25.
.....I know my rocket........
If I remember correctly, people were saying the same thing about a 9 engine cluster around 2008-2010.
Peter,- Engine T/W does not go up as thrust goes down. It just doesn't. Engine T/W tends to be about constant from 100K lbs to 1M lbs thrust class. ISP isn't much effected by thrust either. - You also seem to be obsessed with filling the base of the vehicle with rocket exhaust. You want to find the optimum vehicle thrust to weight (usually around 1.25). Filling the base is not that important.John
Quote from: livingjw on 09/17/2017 07:26 pmPeter,- Engine T/W does not go up as thrust goes down. It just doesn't. Engine T/W tends to be about constant from 100K lbs to 1M lbs thrust class. ISP isn't much effected by thrust either. - You also seem to be obsessed with filling the base of the vehicle with rocket exhaust. You want to find the optimum vehicle thrust to weight (usually around 1.25). Filling the base is not that important.JohnI would have thought that engine TWR would increase with increasing size/thrust although SpX seems to think otherwise. For a given Pc rocket engine the plumbing wall thickness would remain constant so the OD/ID ratio gets smaller as a rocket engine is scaled up so in theory a larger rocket engine should have a higher TWR than a smaller one if all other variables are constant. So a future larger BFR should have the no more than the same no. of engines as mini-BFR. Less engines = less plumbing on the bottom of the BFR.
Quote from: livingjw on 09/17/2017 07:26 pmPeter,.......You also seem to be obsessed with filling the base of the vehicle with rocket exhaust.......Obsessed is right. Too much armchair amateur rocket designing based on personal belief rather than on science, engineering, and technological understanding.
Quote from: DJPledger on 09/17/2017 08:02 pmQuote from: livingjw on 09/17/2017 07:26 pmPeter,- Engine T/W does not go up as thrust goes down. It just doesn't. Engine T/W tends to be about constant from 100K lbs to 1M lbs thrust class. ISP isn't much effected by thrust either. - You also seem to be obsessed with filling the base of the vehicle with rocket exhaust. You want to find the optimum vehicle thrust to weight (usually around 1.25). Filling the base is not that important.JohnI would have thought that engine TWR would increase with increasing size/thrust although SpX seems to think otherwise. For a given Pc rocket engine the plumbing wall thickness would remain constant so the OD/ID ratio gets smaller as a rocket engine is scaled up so in theory a larger rocket engine should have a higher TWR than a smaller one if all other variables are constant. So a future larger BFR should have the no more than the same no. of engines as mini-BFR. Less engines = less plumbing on the bottom of the BFR.No, wall thickness is proportional to diameter. Once you reach a certain minimum size TWR does not change much with thrust.
They've talked about the engine being scaleable -- 'easily' was the adjective used IIRC -- so they will grow whatever size engine(s) they need from the Raptor family. My bet is to see second size at this IAC... sub-scale is likely in or near flight qualification testing.
-snip-
Nice logarithmic chart!T/W for Raptor can't be lower than the 198 for SpaceX own M1D, because otherwise the Raptor wouldn't be the rocket engine with the highest T/W ratio, as it was presented at IAC 2016.If Elon Musk states Raptor has 3 times as much thrust than M1D at a similar size wouldn't that not imply a T/W ratio of around an unimaginably good score of around 500-600?
Quote from: Peter.Colin on 09/18/2017 12:54 amNice logarithmic chart!T/W for Raptor can't be lower than the 198 for SpaceX own M1D, because otherwise the Raptor wouldn't be the rocket engine with the highest T/W ratio, as it was presented at IAC 2016.If Elon Musk states Raptor has 3 times as much thrust than M1D at a similar size wouldn't that not imply a T/W ratio of around an unimaginably good score of around 500-600?Benifits of Full Flow Stage Combustion, in that by adding 50% more turbopump machinery (which is by far not the entirity of the mass of the engine) you can get closed cycle efficencies with a chamber pressure higher than most open cycle engines.
Re scalability: I can't remember where but I read there is a principal limit to chemical rocket engine sizes. Something to do with fluid and combustion dynamics. As the limit is approached, the engine becomes more and more complex. Surely the cycle type and fuel type influences the limit, so raptor is hard to compare with existing engines in this regard. Larger might not be better.
Nice logarithmic chart!T/W for Raptor can't be lower than the 198 for SpaceX own M1D, because otherwise the Raptor wouldn't be the rocket engine with the highest T/W ratio, as it was presented at IAC 2016.If Elon Musk states Raptor has 3 times as much thrust than M1D at a similar size would that not imply a T/W ratio of an unimaginably good score of around 500-600?
Lots of people saying that many engines is bad. Increased chance of RUD, more complexity. And yet SpaceX have flown the 9 engined Falcon 9 with no failures at all for quite a few years. That's a LOT of flight hours on engines with no failures. More complex? No, just more of them, and smaller, which makes removal and inspection easier, and replacement considerably easier. There is quite a bit of plumbing of course, but is that a real issue?So I'm not seeing the problem with large numbers of engines on the stage. Can anyone enlighten as to why it is such a 'bad thing'.
The NK33 utilizes a closed cycle similar to the Raptor. Raptor has two turbines, 2 pumps and 2 pre-burners. NK33 has one turbine, two pumps and one pre-burner. Raptor has higher pressure. Chamber plumbing and pumps scale directly with volume and pressure. The Raptor has the advantage of better materials, analysis, QA, CNC, and 3D printing so you might expect it to have better thrust to weight than the NK33 despite its higher pressure and complexity. T/W of 500-600 for such a design is shear fantasy.John
Quote from: livingjw on 09/18/2017 11:22 amThe NK33 utilizes a closed cycle similar to the Raptor. Raptor has two turbines, 2 pumps and 2 pre-burners. NK33 has one turbine, two pumps and one pre-burner. Raptor has higher pressure. Chamber plumbing and pumps scale directly with volume and pressure. The Raptor has the advantage of better materials, analysis, QA, CNC, and 3D printing so you might expect it to have better thrust to weight than the NK33 despite its higher pressure and complexity. T/W of 500-600 for such a design is shear fantasy.JohnI think the Raptor can get ~200 TWR, I agree 5-600 is fantasy.I think having the oxidizer turbine pretty much integrated into the combustion head looks like a huge saver of weight, along with the close co-location of many of the parts. I imagine high pressure piping is probably a large chunk of the weight of a staged combustion engine. You can see that the design took steps to minimize this as much as possible, Just compare the amount of high pressure piping compared to an RD-170/180. This is an area where CAD/3D printing can have a huge effect compared to 40 years ago.
I have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.John
Quote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnFor anyone interested, I'm willing to make a bet for $50 the Raptor already has at least T/W of above 350.After attaining a high Isp of a rocket engine, attaining a high T/W is the next logical goal.Isp is limited by physics to a theoretical maximum.Nothing in physics is preventing a rocket engine to reach even much higher T/W values.
Quote from: ZachF on 09/18/2017 02:38 pmQuote from: livingjw on 09/18/2017 11:22 amThe NK33 utilizes a closed cycle similar to the Raptor. Raptor has two turbines, 2 pumps and 2 pre-burners. NK33 has one turbine, two pumps and one pre-burner. Raptor has higher pressure. Chamber plumbing and pumps scale directly with volume and pressure. The Raptor has the advantage of better materials, analysis, QA, CNC, and 3D printing so you might expect it to have better thrust to weight than the NK33 despite its higher pressure and complexity. T/W of 500-600 for such a design is shear fantasy.JohnI think the Raptor can get ~200 TWR, I agree 5-600 is fantasy.I think having the oxidizer turbine pretty much integrated into the combustion head looks like a huge saver of weight, along with the close co-location of many of the parts. I imagine high pressure piping is probably a large chunk of the weight of a staged combustion engine. You can see that the design took steps to minimize this as much as possible, Just compare the amount of high pressure piping compared to an RD-170/180. This is an area where CAD/3D printing can have a huge effect compared to 40 years ago.The side by side with the BE-4 on the previous page is pretty amazing too, that is a whole lot of machinery hanging on the side. One can either marvel at the efficiency of the Raptor or the inefficiency of the BE-4.
Quote from: Peter.Colin on 09/18/2017 05:33 pmQuote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnFor anyone interested, I'm willing to make a bet for $50 the Raptor already has at least T/W of above 350.After attaining a high Isp of a rocket engine, attaining a high T/W is the next logical goal.Isp is limited by physics to a theoretical maximum.Nothing in physics is preventing a rocket engine to reach even much higher T/W values.I think it will be hard to get Raptor's TWR much beyond about 200 due to all the turbomachinery and plumbing required for FFSC.I think you will lose your $50 bet.
Quote from: livingjw on 09/18/2017 11:22 amThe NK33 utilizes a closed cycle similar to the Raptor. Raptor has two turbines, 2 pumps and 2 pre-burners. NK33 has one turbine, two pumps and one pre-burner. Raptor has higher pressure. Chamber plumbing and pumps scale directly with volume and pressure. The Raptor has the advantage of better materials, analysis, QA, CNC, and 3D printing so you might expect it to have better thrust to weight than the NK33 despite its higher pressure and complexity. T/W of 500-600 for such a design is shear fantasy.JohnI think you're missing exactly what those two preburners and 2 turbines give you. A single preburner, single turbine closed cycle engine is losing performance to back pressure, because the single preburner is trying to push 3 different flows (fuel, oxydizer, and preburner) into the combustion chamber. In full flow staged combustion, each turbine/preburner is only pushing 1 flow (1 and ahalf, really, since each is also a preburner, assuming equal flow), at least doubling the possible chamber pressure without any other advances.
Quote from: rakaydos on 09/18/2017 05:08 pmQuote from: livingjw on 09/18/2017 11:22 amThe NK33 utilizes a closed cycle similar to the Raptor. Raptor has two turbines, 2 pumps and 2 pre-burners. NK33 has one turbine, two pumps and one pre-burner. Raptor has higher pressure. Chamber plumbing and pumps scale directly with volume and pressure. The Raptor has the advantage of better materials, analysis, QA, CNC, and 3D printing so you might expect it to have better thrust to weight than the NK33 despite its higher pressure and complexity. T/W of 500-600 for such a design is shear fantasy.JohnI think you're missing exactly what those two preburners and 2 turbines give you. A single preburner, single turbine closed cycle engine is losing performance to back pressure, because the single preburner is trying to push 3 different flows (fuel, oxydizer, and preburner) into the combustion chamber. In full flow staged combustion, each turbine/preburner is only pushing 1 flow (1 and ahalf, really, since each is also a preburner, assuming equal flow), at least doubling the possible chamber pressure without any other advances.While FFSC was always considered harder than ORSC, I wouldn't be surprised if that is only because it was harder with the engineering tools of a few decades ago, but may no longer be the case. Raptor seems to be going along pretty smoothly while BE-4 and AR-1 seem a bit more stuck.Simulating the complex startup and flows of FFSC might have been extremely hard with 70s technology, but much easier with today's advanced computers. Meanwhile, the materials problems that arise from ORSC haven't changed much.I would not be surprised to hear Raptor is progressing better than expected at IAC.
Quote from: JamesH65 on 09/18/2017 12:17 pmLots of people saying that many engines is bad. Increased chance of RUD, more complexity. And yet SpaceX have flown the 9 engined Falcon 9 with no failures at all for quite a few years. That's a LOT of flight hours on engines with no failures. More complex? No, just more of them, and smaller, which makes removal and inspection easier, and replacement considerably easier. There is quite a bit of plumbing of course, but is that a real issue?So I'm not seeing the problem with large numbers of engines on the stage. Can anyone enlighten as to why it is such a 'bad thing'.7-9 is the optimal no. for a 1st stage which is why NG and F9 have these engine nos. It's when you go beyond about 20 (look what happened to the N-1) on the 1st stage that you are likely to enter problems with increased risk of RUD's causing LOM, higher maintenance costs and increased downtime of maintaining all those engines. If SpaceX ever builds a c.120-130MN thrust booster they should go with a scaled up Raptor for it. Lower risk of LOM coupled with lower maintenance costs and less downtime between missions may outweigh a slight reduction in engine TWR. Just make the booster slightly larger to compensate and use the TWR optimized size Raptors for the ITS ship which needs the highest TWR and performance engines.
Quote from: DJPledger on 09/18/2017 05:45 pmQuote from: JamesH65 on 09/18/2017 12:17 pmLots of people saying that many engines is bad. Increased chance of RUD, more complexity. And yet SpaceX have flown the 9 engined Falcon 9 with no failures at all for quite a few years. That's a LOT of flight hours on engines with no failures. More complex? No, just more of them, and smaller, which makes removal and inspection easier, and replacement considerably easier. There is quite a bit of plumbing of course, but is that a real issue?So I'm not seeing the problem with large numbers of engines on the stage. Can anyone enlighten as to why it is such a 'bad thing'.7-9 is the optimal no. for a 1st stage which is why NG and F9 have these engine nos. It's when you go beyond about 20 (look what happened to the N-1) on the 1st stage that you are likely to enter problems with increased risk of RUD's causing LOM, higher maintenance costs and increased downtime of maintaining all those engines. If SpaceX ever builds a c.120-130MN thrust booster they should go with a scaled up Raptor for it. Lower risk of LOM coupled with lower maintenance costs and less downtime between missions may outweigh a slight reduction in engine TWR. Just make the booster slightly larger to compensate and use the TWR optimized size Raptors for the ITS ship which needs the highest TWR and performance engines.7-9 engines on the first stage is only optimal if the first stage is reusable. If you want to have common engines (sea level and vacuum versions) on both stages, you need more than 9 on the first stage.Original ITS:1st stage: 42 atmospheric engines2nd stage: 6 vacuum + 3 atmospheric enginesHalve the number of engines:1st stage: 21 atmospheric engines2nd stage: 3 vacuum + 1-2 atmospheric enginesIf there were 9 engines on the first stage, you would need a separate engine production line for the second stage in order to have a manageable T/W ratio or throttling capability for landing.
Looks like mini-BFR will have 19-21 engines on booster because SpX are making the Raptor too small for 9 engines to generate sufficient thrust. We will find out soon at IAC2017. The original plan for BFR was for 9 engines in the F-1 class so I don't understand why SpX are going for so many small engines. 19-21 engines on booster may end up being acceptable for all we know but future larger BFR's should not go for any more engines than this. Perhaps BO will be more sensible with the engine nos. than SpX for their future HLV's.
Quote from: DJPledger on 09/18/2017 07:21 pmLooks like mini-BFR will have 19-21 engines on booster because SpX are making the Raptor too small for 9 engines to generate sufficient thrust. We will find out soon at IAC2017. The original plan for BFR was for 9 engines in the F-1 class so I don't understand why SpX are going for so many small engines. 19-21 engines on booster may end up being acceptable for all we know but future larger BFR's should not go for any more engines than this. Perhaps BO will be more sensible with the engine nos. than SpX for their future HLV's.They went for many small engines because they discovered that "optimum number of engines" was a lot more than they originally thought.
Quote from: Pipcard on 09/18/2017 06:44 pmQuote from: DJPledger on 09/18/2017 05:45 pmQuote from: JamesH65 on 09/18/2017 12:17 pmLots of people saying that many engines is bad. Increased chance of RUD, more complexity. And yet SpaceX have flown the 9 engined Falcon 9 with no failures at all for quite a few years. That's a LOT of flight hours on engines with no failures. More complex? No, just more of them, and smaller, which makes removal and inspection easier, and replacement considerably easier. There is quite a bit of plumbing of course, but is that a real issue?So I'm not seeing the problem with large numbers of engines on the stage. Can anyone enlighten as to why it is such a 'bad thing'.7-9 is the optimal no. for a 1st stage which is why NG and F9 have these engine nos. It's when you go beyond about 20 (look what happened to the N-1) on the 1st stage that you are likely to enter problems with increased risk of RUD's causing LOM, higher maintenance costs and increased downtime of maintaining all those engines. If SpaceX ever builds a c.120-130MN thrust booster they should go with a scaled up Raptor for it. Lower risk of LOM coupled with lower maintenance costs and less downtime between missions may outweigh a slight reduction in engine TWR. Just make the booster slightly larger to compensate and use the TWR optimized size Raptors for the ITS ship which needs the highest TWR and performance engines.7-9 engines on the first stage is only optimal if the first stage is reusable. If you want to have common engines (sea level and vacuum versions) on both stages, you need more than 9 on the first stage.Original ITS:1st stage: 42 atmospheric engines2nd stage: 6 vacuum + 3 atmospheric enginesHalve the number of engines:1st stage: 21 atmospheric engines2nd stage: 3 vacuum + 1-2 atmospheric enginesIf there were 9 engines on the first stage, you would need a separate engine production line for the second stage in order to have a manageable T/W ratio or throttling capability for landing.The 1st stage is reusable on both NG and F9 so have optimum engine nos. and have common US engines. Having two separate engine production lines for two sizes of the same fundamental design is not a big deal these days with modern manufacturing methods. Larger Raptors for BFR booster and smaller Raptors for ITS ship to keep engine no. on booster to around 7-9 and the same engine no. on ship. Or make Raptor so deeply throttlable that you can have a single engine on the ship so as to keep optimum engine no. on booster while keeping only one engine production line.Looks like mini-BFR will have 19-21 engines on booster because SpX are making the Raptor too small for 9 engines to generate sufficient thrust. We will find out soon at IAC2017. The original plan for BFR was for 9 engines in the F-1 class so I don't understand why SpX are going for so many small engines. 19-21 engines on booster may end up being acceptable for all we know but future larger BFR's should not go for any more engines than this. Perhaps BO will be more sensible with the engine nos. than SpX for their future HLV's.
Quote from: Peter.Colin on 09/18/2017 05:33 pmQuote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnFor anyone interested, I'm willing to make a bet for $50 the Raptor already has at least T/W of above 350.After attaining a high Isp of a rocket engine, attaining a high T/W is the next logical goal.Isp is limited by physics to a theoretical maximum.Nothing in physics is preventing a rocket engine to reach even much higher T/W values.Peter, I agree and Raptor will most likely dethrone the NK33 which is currently the highest thrust to weight staged combustion rocket that exists. I have my doubts, but it may top 200, but I'm sticking with around 160. The chemistry and thermodynamics of the Raptor cycle mostly known. Material properties and efficient mechanical design will ultimately dictate T/W. Material Properties are a known commodity as well. They have improved since the 1970s but not by so much that you could get anywhere near 350. I can assure you that the Raptor is wringing every bit of strength to weight available out of its materials commensurate with reliability and life requirements. Rocket designers do that. This may help: The M1D weights about 1000 lbs and has a chamber pressure of about 10-MPa. The Raptor has a chamber pressure of 30-MPa and pre-burner pressures somewhere around 45-MPa. Pumps, plumbing, valves, pre-burners and the main combustion chamber all scale pretty much directly with pressure, volume and material strength to weight. The M1D is about the same size as the Raptor so lets just scale its weight up by the difference in pressure between the two engines. (I'm ignoring the higher pump and pre-burner pressure!) That alone gives you at least 3000 lbs weight and a thrust to weight of 220. We still haven't accounted for the higher pressure pumps and pre-burners, or the fact that the Raptor has an additional pre-burner and turbine. If we add an additional 400-1000 lbs to account for these weights we are in the 160 = 190 T/W ball park. In summary, without some new wonder materials, 350 T/W is not in the cards.
OK agreed to the bet, I hope I'm not wrong for multiple reasons...
Quote from: Peter.Colin on 09/18/2017 11:21 pmOK agreed to the bet, I hope I'm not wrong for multiple reasons... Pro tip, Peter... Don't go to Vegas. You have already made one bet with hideous odds, why are you itching to lose money by more?
Quote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnJohn,I'm curious, any idea how subcooling propellants may affect the T/W ratio?Steven
Quote from: wannamoonbase on 09/19/2017 01:13 pmQuote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnJohn,I'm curious, any idea how subcooling propellants may affect the T/W ratio?StevenIt doesn't, really. The difference between subcooled and boiling is negligible compared to the temperature in the combustion chamber. As long as the turbopumps can maintain the same chamber pressure, thrust will be about the same.
Quote from: envy887 on 09/19/2017 02:10 pmQuote from: wannamoonbase on 09/19/2017 01:13 pmQuote from: livingjw on 09/17/2017 11:57 pmI have an old chart from K. D. Wood's spacecraft Design book that shows the general trend for rocket engine T/Ws.It is a bit dated, but so are most rocket engines. This chart shows that thrust to weights are nearly flat between 50 klbs and 1 mlbs. I have spotted the M1D and NK33. I would expect the Raptor T/W to be somewhere between these two. Lets guess T/W = 160. I think OneSpeed's thrust guess at 2.5 mN sounds about right. The improvement in SpaceX's T/Ws comes from improved material, analysis, QA, accurate CNC and 3D printing technologies. I can safely say that the chemistry and thermodynamics have not changed one bit since this chart was made. There is nothing in the Raptor chemistry or design which would allow it to deviate from normal sizing trends.JohnJohn,I'm curious, any idea how subcooling propellants may affect the T/W ratio?StevenIt doesn't, really. The difference between subcooled and boiling is negligible compared to the temperature in the combustion chamber. As long as the turbopumps can maintain the same chamber pressure, thrust will be about the same.So it has no influence on Thrust, but does it have on weight? Maybe the turbopump becomes more efficient or the cooling lines around the engine chamber can take more energy so that the chamber can be operated at higher temperatures? I dont know the answer, I am asking.
So a sphere 2x bigger at the same pressure is 8x heavier?The way I figure it is 2x the diameter is 2x longer circumference which means 2x greater force which means 2x thicker walls.Now if the walls of a sphere are 2x thicker than the whole surface area of the sphere goes as x^2 so the mass goes as x^3So to summarize:1. the weight of an engine goes as the cube of scaling.x^32x size increase = 8x weightThis is because the combustion chamber, rocket nozzle, and plumbing are all like a pressure vessel and go as the cube.Since thrust goes as the throat area than the thrust is x^2So smaller engines win out on weight.2. If the pressure is doubled then the thrust doubles, and the wall thickness doubles, and the weight doubles.So no difference on weight per thrust.Do I have these heuristics right?All of this is assuming that the major weight of the engine is in the pressure components.Obviously there are other components and they may not scale the same.
Quote from: rsdavis9 on 09/19/2017 03:44 pmSo a sphere 2x bigger at the same pressure is 8x heavier?The way I figure it is 2x the diameter is 2x longer circumference which means 2x greater force which means 2x thicker walls.Now if the walls of a sphere are 2x thicker than the whole surface area of the sphere goes as x^2 so the mass goes as x^3So to summarize:1. the weight of an engine goes as the cube of scaling.x^32x size increase = 8x weightThis is because the combustion chamber, rocket nozzle, and plumbing are all like a pressure vessel and go as the cube.Since thrust goes as the throat area than the thrust is x^2So smaller engines win out on weight.2. If the pressure is doubled then the thrust doubles, and the wall thickness doubles, and the weight doubles.So no difference on weight per thrust.Do I have these heuristics right?All of this is assuming that the major weight of the engine is in the pressure components.Obviously there are other components and they may not scale the same.You are mostly right except:2. If pressure is doubled then wall thickness is less than doubled so the weight is also less than doubled.So higher pressure at same engine dimensions also leads to higher T/W ratio as do smaller engines (provided their Isp remains high if you shrink them).
You are mostly right except:2. If pressure is doubled then wall thickness is less than doubled so the weight is also less than doubled.So higher pressure at same engine dimensions also leads to higher T/W ratio as do smaller engines (provided their Isp remains high if you shrink them).
Quote from: Peter.Colin on 09/19/2017 04:20 pmYou are mostly right except:2. If pressure is doubled then wall thickness is less than doubled so the weight is also less than doubled.So higher pressure at same engine dimensions also leads to higher T/W ratio as do smaller engines (provided their Isp remains high if you shrink them).Ok I give up. Why is wall thickness less than doubled?Under pressure(tension) it should be exact.Under compression you have buckling and other things.
...A double chamber pressure does not lead to double pressure at the end of the expansion nozzle cone.So the end of the cone does not have to be twice as thick only the beginning of the cone.
Because the cone has the largest contribution to the weight I suspect higher chamber pressure is usually positive for T/W.
...Most of the mass is in the main combustion chamber, injector head, turbopumps, and high pressure plumbing.
Quote from: Peter.Colin on 09/19/2017 06:16 pm...A double chamber pressure does not lead to double pressure at the end of the expansion nozzle cone.So the end of the cone does not have to be twice as thick only the beginning of the cone.Most nozzles are not, strictly speaking, cones. They are bell-shapes.Your statement is only true of the net pressure when operating in the atmosphere; but all rocket engines operate in vacuum eventually, where the pressure drop across the nozzle is nearly linear with chamber pressure.QuoteBecause the cone has the largest contribution to the weight I suspect higher chamber pressure is usually positive for T/W.The effect you are basing this on is flawed, and also, the nozzle is actually a small part of the weight. For instance, the SSME had a large nozzle for it's thrust (70:1 expansion, full LH2 regen cooling), but it was still only about 12% of the total engine mass.Most of the mass is in the main combustion chamber, injector head, turbopumps, and high pressure plumbing.
Quote from: envy887 on 09/19/2017 06:47 pm...Most of the mass is in the main combustion chamber, injector head, turbopumps, and high pressure plumbing.I think some people are dismissing the pressure. 300 bar, is no joke.I've worked with carbon steel and stainless steel in 5000-7000 psi range for heat exchanges and piping. It's no joke.When the first Raptor info came out last year stating 300 bar I thought that was the most impressive number. Holding everything together at 4350 psi with cryogenics coming in one end and rocket exhaust going out the other is no minor task.
rsdrvis9, Close, but not quite: - The combustion physics requires a certain dwell time in order to efficiently mix and burn. This dwell time is captured by the parameter "Characteristic Length", L* = Vc/Ath. RP1/Lox L* = 102-127 cm. CH4/Lox is probably a little higher maybe around 150-160 cm. So knowing your chemistry gives you an L* which allows you to calculate a combustion chamber volume, Vc = L* x Ath; hence, Vc is proportional to Ath. If you look at rockets of different thrust, you will clearly see that lower thrust rockets have proportionally bigger combustion chambers and higher thrust rockets have smaller chambers. The same goes for the pre-burners. Also note that these components handle the highest pressures. To summarize: - For a given cycle, a rocket engine weight scales roughly with thrust because: - Throat area and mass flow are proportional to thrust. - The pre-burners, combustion chamber and plumbing are pressure vessels proportional to throat area. - The turbines and pumps are proportional to pressure and mass flow. - Most of the expansion nozzle contains low pressure, its weight is dominated by minimal unit wall weight proportional to nozzle area.- A sphere 2x bigger has 4x the surface area and 8x the volume. At the same pressure,with the same material allowables it is 8x heavier, so the weight per unit volume stays the same. This is classic pressure vessel behavior.John
Quote from: livingjw on 09/19/2017 07:20 pmrsdrvis9, Close, but not quite: - The combustion physics requires a certain dwell time in order to efficiently mix and burn. This dwell time is captured by the parameter "Characteristic Length", L* = Vc/Ath. RP1/Lox L* = 102-127 cm. CH4/Lox is probably a little higher maybe around 150-160 cm. So knowing your chemistry gives you an L* which allows you to calculate a combustion chamber volume, Vc = L* x Ath; hence, Vc is proportional to Ath. If you look at rockets of different thrust, you will clearly see that lower thrust rockets have proportionally bigger combustion chambers and higher thrust rockets have smaller chambers. The same goes for the pre-burners. Also note that these components handle the highest pressures. To summarize: - For a given cycle, a rocket engine weight scales roughly with thrust because: - Throat area and mass flow are proportional to thrust. - The pre-burners, combustion chamber and plumbing are pressure vessels proportional to throat area. - The turbines and pumps are proportional to pressure and mass flow. - Most of the expansion nozzle contains low pressure, its weight is dominated by minimal unit wall weight proportional to nozzle area.- A sphere 2x bigger has 4x the surface area and 8x the volume. At the same pressure,with the same material allowables it is 8x heavier, so the weight per unit volume stays the same. This is classic pressure vessel behavior.JohnJohn, your analogy doesnt work because Raptor has both gaseous Oxygen and gaseous Methane mixing. I bet that the combustion physics of Raptor is unlike any other engine because of this. It may very well be that this leads to a very high T/W ratio. At least more than you would expect in a liquid-liquid engine.
Quote from: Semmel on 09/19/2017 07:25 pmJohn, your analogy doesnt work because Raptor has both gaseous Oxygen and gaseous Methane mixing. I bet that the combustion physics of Raptor is unlike any other engine because of this. It may very well be that this leads to a very high T/W ratio. At least more than you would expect in a liquid-liquid engine.It is not an analogy. Its engineering. L* does depend on mixing details and may be lower for gas/gas mixing in the main combustion chamber. That does not invalidate the engineering approach.
John, your analogy doesnt work because Raptor has both gaseous Oxygen and gaseous Methane mixing. I bet that the combustion physics of Raptor is unlike any other engine because of this. It may very well be that this leads to a very high T/W ratio. At least more than you would expect in a liquid-liquid engine.
Quote from: livingjw on 09/19/2017 07:34 pmQuote from: Semmel on 09/19/2017 07:25 pmJohn, your analogy doesnt work because Raptor has both gaseous Oxygen and gaseous Methane mixing. I bet that the combustion physics of Raptor is unlike any other engine because of this. It may very well be that this leads to a very high T/W ratio. At least more than you would expect in a liquid-liquid engine.It is not an analogy. Its engineering. L* does depend on mixing details and may be lower for gas/gas mixing in the main combustion chamber. That does not invalidate the engineering approach.Sorry, didnt mean to be emphasize the term 'analogy'. I enjoy your engineering approach (not only here) quite a bit. I however wanted to point out that your point about the length of the cumbustion chamber actually should be reversed. Hence a higher T/W ratio and the balancing factors do not apply here.Also, as ZachF pointed out, integrating the LOX preburner/turbine onto the head of the injector makes quite a lot of sense. I bet this design saved them quite a lot of grief.
- My post addressed why T/W tends to stay constant when you scale a rocket engine. I agree that integrating the LOX pre-burner/turbine into the head of the main chamber should increase T/W. I also agree that gas/gas mixing and combustion might have a lower L* than liquid/liquid or liquid/gas, which would also increase T/W. These are the types of details that can get you from 160 - 200 T/W. They won't get you to 350. Either way, the T/W trend with thrust will continue to be relatively flat.
Out of interest, can you suggest a good book on rocket engine engineering? I am not afraid of math.
Quote from: Semmel on 09/19/2017 09:31 pmOut of interest, can you suggest a good book on rocket engine engineering? I am not afraid of math.Rocket Propulsion Elements - SuttonMy copy is about 20 years old, but the 9th edition is Feb 2017.
Quote from: MikeAtkinson on 09/19/2017 10:14 pmQuote from: Semmel on 09/19/2017 09:31 pmOut of interest, can you suggest a good book on rocket engine engineering? I am not afraid of math.Rocket Propulsion Elements - SuttonMy copy is about 20 years old, but the 9th edition is Feb 2017.NASA SP-125 is still one of the most complete. I have attached the PDF (Don't you love the internet!). It covers theory to practical design details with lots of drawings and graphs. Couple this with NASA's online CEA program and you have a very good start. SP-125 mostly covers gas generator cycles. Your education won't be complete until you dig into the Russian staged combustion engines. They are truly phenomenal. In SP-125 a staged combustion cycle is referred to as dual combustion cycle. I had Sutton's book for an undergraduate course in 1971. It was OK, but not as much detail as in SP-125. SP-125 helped get us to the Moon.
I had the impresssion occasionally that developing rocket engines was a lost art in the US, reinvented only when SpaceX and BO appeared on the scene.So many threads on potential new launch vehicles over the years and they always looked which engines are available off the shelf, not developing a new engine tailormade for the needs.
Peter,Yes, they are doing very good work. More to the point, SpaceX is modeling the mixing and combustion chemistry. Combustion is mostly limited by mixing, and mixing as the video shows is fractal crazy! This kind of work goes back decades, but is now advancing rapidly with the advent of affordable massively parallel computers. The combustion reactions themselves go back even further and are well characterized. I just now went to NASA's old Chemical Equilibrium Analysis, CEA sight and ran a Methane-Oxygen case. See attached. The chemical equilibrium assumption works very well for rocket chambers, not so well for nozzles, but we can approximate non-equilibrium effects well enough. I doubt we will see any surprises, just lots of great detail to guide the design and development.John
Where can I get that "Nozzle Pack" software? Google isn't being particularly helpful in this case.
Things can be done in parallel. You don't have to wait for the engines to pass qualification before you design the rest of the booster.
Has anybody considered that instead of individual TVC for each engine that the whole rocket could be steered with differential thrust?So for example with the falcon 9 get rid of TVC and use differential thrust on different engines around the ring of engines on the outside of the rocket.Still might have a problem with roll but that could be solved by making the outer ring slightly pointed right then left as you go around the ring and just applying different thrust to the odd or even engines.For landing you probably want one engine with TVC.It could save a lot of weight.With lots of engines I would expect finer control of the vector.Should be a greater vector with larger diameter rockets.
Quote from: rsdavis9 on 09/21/2017 02:47 pmHas anybody considered that instead of individual TVC for each engine that the whole rocket could be steered with differential thrust?So for example with the falcon 9 get rid of TVC and use differential thrust on different engines around the ring of engines on the outside of the rocket.Still might have a problem with roll but that could be solved by making the outer ring slightly pointed right then left as you go around the ring and just applying different thrust to the odd or even engines.For landing you probably want one engine with TVC.It could save a lot of weight.With lots of engines I would expect finer control of the vector.Should be a greater vector with larger diameter rockets.Last year's ITS booster was presented with rings of differential thrusting non-gimballing engines surrounding a TVC gimballed central cluster. Dragon 2 uses differential thrust steering for the abort motors. And N-1's booster used differential thrust for steering.
Thinking about the proportionality of rockets and rocket engines made me wonder about the following: Would it have been easier to colonize mars if the human species was smaller than it is now, or bigger than it is now?Or are we ourselves as a humans at 1,80 meter coincidently at the optimal size to build similar sized rockets engines and become a space fairing civilization?If we where smaller we might have built smaller rocket engines with higher T/W or build the exact same “optimal” engine size as we do now, but less of them where needed, to get 100 people to Mars. Or if we would have been bigger than now, we would be able to build large engines more easily, but maybe get in trouble with material properties not being sufficient. The only option then would be to get many small engines on a large rocket.And what size would an alien space fairing civilization be, if compared to our own size?
Quote from: jpo234 on 09/21/2017 09:36 amThings can be done in parallel. You don't have to wait for the engines to pass qualification before you design the rest of the booster.True but the engine needs to have completed some test fires. In case of full scale Raptor we've not heard of any being tested.
Quote from: Peter.Colin on 09/21/2017 05:51 pmThinking about the proportionality of rockets and rocket engines made me wonder about the following: Would it have been easier to colonize mars if the human species was smaller than it is now, or bigger than it is now?Or are we ourselves as a humans at 1,80 meter coincidently at the optimal size to build similar sized rockets engines and become a space fairing civilization?If we where smaller we might have built smaller rocket engines with higher T/W or build the exact same “optimal” engine size as we do now, but less of them where needed, to get 100 people to Mars. Or if we would have been bigger than now, we would be able to build large engines more easily, but maybe get in trouble with material properties not being sufficient. The only option then would be to get many small engines on a large rocket.And what size would an alien space fairing civilization be, if compared to our own size?Throttle levels are much slower to respond and less precise than hydraulic TVC. Whether differential throttling could be fast or precise enough I think is unlikely.
After seeing the presentation today, I feel like the new size and new design of the BFR makes it a perfect multi utility workhorse.It could be in service for a long time before the next BFR will be realized.What size would make this one seem like a rowboat, as Musk said the 12 meter would seem in the future.And is a bigger size even nescerry?
Quote from: Peter.Colin on 09/29/2017 10:33 pmAfter seeing the presentation today, I feel like the new size and new design of the BFR makes it a perfect multi utility workhorse.It could be in service for a long time before the next BFR will be realized.What size would make this one seem like a rowboat, as Musk said the 12 meter would seem in the future.And is a bigger size even nescerry?I think that if this BFR actually gets built and flies that it's the larger BFR doesn't get built. A 12 million Lbf rocket, is not a small rocket, this could do a ton of work.