Is it possible to create Saturn V(esqe) rocket but with entirely modern/newer technology? What its performance will be like?

[magnemoe]

56 minutes ago, Reactordrone said:

RD-170s have similar thrust to an F-1 so you could do a first stage without external boosters and carry 100 tonnes less propellant for the same delta-v. That would also reduce gravity drag since your TWR would be higher (and the Saturn V had terrible TWR on the pad).

True, however the trend before reuse was SRB as they are cheaper than rocket engines.
Low TWR also made sense as fuel and its tanks are cheaper than engines. 
Starships superheavy on the other hand will be high TWR as it want to get starship up to an suborbital trajectory fast so it can do it boostback early. 

[sevenperforce]

14 hours ago, RCgothic said:

I guess that makes sense! I'd just never heard it described that way before!

I hadn't either, not until a few months ago, but it REALLY simplifies calculations. Especially when you are working with real flight telemetry. No more piecewise integration.

Quote

I think your Merlin 1D figures are wrong though. 854kN thrust according to Wikipedia.

Yep, typo. Thanks. The actual number is 942 kN because I am using vac thrust for all engines.

Edited April 25, 2020 by sevenperforce

[sevenperforce]

I went back and used actual numbers from the Apollo 17 mission Ascent Data to determine the real performance parameters for Saturn V.

Drag likely includes propellant residuals. All values in m/s.

Stage	Onboard dV	Drag Losses	Effective dV
S-IVB (TLI)	3122.0	82.1	3039.9
S-IVB (EOI)	1022.1	210.3	811.8
S-II	4746.9	547.4	4199.4
S-IC	3869.4	1087.6	2781.8
Vehicle	12760.3	1927.4	10832.9

Total payload to TLI was 48.6 tonnes.

S-II was a friggin' workhorse.

Edited April 25, 2020 by sevenperforce

[sevenperforce]

More engine data...

Engines						bulk density (kg/m^3)
Engine	Thrust (kN)	Weight (kg)	isp_v (s)	prop mass, kg	O/F	ox	fuel	prop at mix
Merlin 1D	942	470	311	-	2.34	1252	826	1124.3
RD-180	4152	519	338	-	2.60	1141	810	1049.1
RL-10C-1	101.9	301	449.7	-	5.50	1141	709	1074.5
RL-10C-1-1	105.9	301	453.8	-	5.50	1141	709	1074.5
RL-10C-2	111.2	301	465.5	-	5.50	1141	709	1074.5
RL-10C-3	108.5	301	460.1	-	5.50	1141	709	1074.5
Raptor Vac	2150	1826	380	-	3.80	1252	439	1082.3
Raptor SL	1960	1500	350	-	3.80	1252	439	1082.3
BE-4 SL	2400	5246	340	-	3.70	1141	423	988.1
BE-3U	710	1477	452	-	5.50	1141	709	1074.5
Merlin 1D Vac	981	480	380	-	2.36	1252	826	1125.0
RS-25 (SSME)	2279	3526	452.3	-	6.00	1141	709	1079.2
RS-68A	3556	6686	411	-	6.00	1141	709	1079.2
SLS SRB	20000	113750	268	625000	-	-	-	-
GEM-63XL	2026	5400	284	48000	-	-	-	-

Interesting that the first stage of the New Glenn has the worst possible bulk prop density.

I didn't realize how much hydrolox engines benefited from that high O/F ratio.

Edited April 25, 2020 by sevenperforce

[magnemoe]

15 hours ago, Reactordrone said:

RD-170s have similar thrust to an F-1 so you could do a first stage without external boosters and carry 100 tonnes less propellant for the same delta-v. That would also reduce gravity drag since your TWR would be higher (and the Saturn V had terrible TWR on the pad).

Saturn 5 had an low TWR as fuel and larger tanks are cheap compared to engines so its an cheap way to increase you payload. Saturn 5 was designed for the moon missions after all. 
This is taken to the extreme with SRB because they are cheap but give lots of trust so you have core with an TWR less than one but add lots of SRB trust. 

Now one easy way to improve Saturn 5 would be to give second and 3rd stage vacuum nozzles. 

[sevenperforce]

22 minutes ago, magnemoe said:

Now one easy way to improve Saturn 5 would be to give second and 3rd stage vacuum nozzles. 

What do you mean? The J-2 absolutely had a vacuum nozzle.

[Reactordrone]

13 hours ago, sevenperforce said:

More engine data...

Engines						bulk density (kg/m^3)
Engine	Thrust (kN)	Weight (kg)	isp_v (s)	prop mass, kg	O/F	ox	fuel	prop at mix
Merlin 1D	942	470	311	-	2.34	1252	826	1124.3
RD-180	4152	519	338	-	2.60	1141	810	1049.1
RL-10C-1	101.9	301	449.7	-	5.50	1141	709	1074.5
RL-10C-1-1	105.9	301	453.8	-	5.50	1141	709	1074.5
RL-10C-2	111.2	301	465.5	-	5.50	1141	709	1074.5
RL-10C-3	108.5	301	460.1	-	5.50	1141	709	1074.5
Raptor Vac	2150	1826	380	-	3.80	1252	439	1082.3
Raptor SL	1960	1500	350	-	3.80	1252	439	1082.3
BE-4 SL	2400	5246	340	-	3.70	1141	423	988.1
BE-3U	710	1477	452	-	5.50	1141	709	1074.5
Merlin 1D Vac	981	480	380	-	2.36	1252	826	1125.0
RS-25 (SSME)	2279	3526	452.3	-	6.00	1141	709	1079.2
RS-68A	3556	6686	411	-	6.00	1141	709	1079.2
SLS SRB	20000	113750	268	625000	-	-	-	-
GEM-63XL	2026	5400	284	48000	-	-	-	-

Interesting that the first stage of the New Glenn has the worst possible bulk prop density.

I didn't realize how much hydrolox engines benefited from that high O/F ratio.

That liquid hydrogen density looks off by a factor of 10. Should be 70.9

RS-25 and RS-68 will end up the same as the BE-4 SL at 988.1, the RL-10s get slightly less prop density at 976.4

Edited April 26, 2020 by Reactordrone
Added corrected figures

[sevenperforce]

28 minutes ago, Reactordrone said:

That liquid hydrogen density looks off by a factor of 10. Should be 70.9

RS-25 and RS-68 will end up the same as the BE-4 SL at 988.1, the RL-10s get slightly less prop density at 976.4

I actually messed it up even worse than that; I wrote the equation to take the mixture ratios as the volumetric ratios so it was all out of whack.

BE-4 is methalox.

Here's the corrected table.

Engines						bulk density (kg/m^3)
Engine	Thrust (kN)	Weight (kg)	isp_v (s)	Length (m)	Dia (m)	prop mass, kg	O/F	ox	fuel	prop at mix
Merlin 1D	914.1	470	311	-	0.91	-	2.34	1252	826	1084.5
RD-180	4152.0	5480	338	-	3.15	-	2.60	1141	810	1024.7
RL-10B-2-1	111.2	301	465.5	2.19	2.14	-	5.88	1141	71	357.6
RL-10C-1	101.9	150.5	449.7	2.19	1.45	-	5.50	1141	71	343.8
RL-10C-1-1	105.9	150.5	453.8	2.46	1.57	-	5.50	1141	71	343.8
RL-10C-2	111.2	301	465.5	2.19	1.86	-	5.50	1141	71	343.8
RL-10C-3	108.5	225.75	460.1	2.19	2.14	-	5.50	1141	71	343.8
Raptor Vac	2150.0	1826	380	6.60	5.81	-	3.80	1252	439	903.2
Raptor SL	1960.0	1500	350	-	1.30	-	3.80	1252	439	903.2
BE-4 SL	2400.0	5246	340	-	1.90	-	3.70	1141	423	837.9
BE-3U	710.0	1477	452	3.18	1.82	-	5.50	1141	71	343.8
Merlin 1D Vac	934.0	480	380	4.82	2.89	-	2.36	1252	826	1085.3
RS-25 (SSME)	2279.0	3526	452.3	-	2.44	-	6.00	1141	71	361.9
RS-68A	3556.0	6686	411	-	2.43	-	6.04	1141	71	363.3
SLS SRB	20000	113750	268	-	-	625000	-	-	-	-
GEM-63XL	2026	5400	284	-	-	48000	-	-	-	-

Hydrolox doesn't even break 400 kg/m³.

 

Edited April 26, 2020 by sevenperforce

[wumpus]

On 4/25/2020 at 10:01 AM, sevenperforce said:

Hydrolox doesn't even break 400 kg/m³.

Pretty scary how much room it takes when it still needs twice as much oxygen as kerolox...

(quoted time incorrect as I can't delete the empty table).

[farmerben]

First stage TWR was a measly 1.15

 

Most KSP players want TWR between 1.5 - 2.  And have flown with TWR's well above and below that, and know what it's like.    

[magnemoe]

On 4/25/2020 at 8:08 PM, sevenperforce said:

What do you mean? The J-2 absolutely had a vacuum nozzle.

It just looks small compared to other vacuum engines, on the other hand the Saturn 5 was massive and the lower trust compared to fist stage fools me. 

32 minutes ago, farmerben said:

First stage TWR was a measly 1.15

Most KSP players want TWR between 1.5 - 2.  And have flown with TWR's well above and below that, and know what it's like.    

The engines was very expensive, extending fuel tanks and adding more fuel is cheap. 
Flight computer was designed with the low TWR in mind. 

[Bill Phil]

19 minutes ago, magnemoe said:

It just looks small compared to other vacuum engines, on the other hand the Saturn 5 was massive and the lower trust compared to fist stage fools me. 

The J-2’s expansion ratio is much smaller than, say, an RL10 or even the SSME. As a hydrolox engine its performance is actually pretty low. The HG-3 developed from the J-2 improved performance significantly, and was a precursor to the later SSME.

[sevenperforce]

As could have been predicted from some of my earlier posts, I spent entirely too much time this weekend setting up a spreadsheet that would let me play rocket legos with a three-stage moon launcher. It's actually quite detailed; it automatically accounts for the various sources of drag based on the chosen parameters and gives you T/W ratios and burn times for each stage to make sure you aren't building something that can't actually fly.

The spreadsheet lets you choose a payload mass, an overall vehicle diameter, and whether you want a payload fairing, a launch abort system, and so forth. Then, for each stage, you select the engine you want, choose how many engines you plan on clustering, and the total height of your stage. You also have to choose what I call "structural density", a rough measure of the weight of tanks at a given length and diameter, but I use exemplary figures from the Delta IV, SLS, Atlas V, Soyuz-2, and so forth to constrain within reasonable limits. It adds the weight of each interstage based on the length of the engine you choose (and adjusts for engines with extensible skirts). Default is a constant tank diameter but it can be manually changed. It will alert you if your first stage burns out too low, if your second stage is oversized and reaches orbit on its own, or if your fineness ratios are out of envelope.

I created a "sandbox" version of the spreadsheet here if people want to play around with it: https://docs.google.com/spreadsheets/d/1UNuPfjZE-YVIrmLINiUMcOzJk-voZ2TqM6R7MgyaM-I/edit?usp=sharing

If you want to change the other parameters you'll need to copy it into your own drive and change the permissions.

Playing around with numbers produced some interesting results:

• A simple Saturn V clone with the same tank sizes (but with modern construction) would throw 53 tonnes to TLI (4.4 tonnes more than Apollo 17) with 7 RD-180s on the first stage, 7 BE-3Us on the second stage, and 3 BE-3Us on the third stage.
• Swapping out BE-3Us for the RL-10C-1-1 is not a good idea. They have so little thrust that even if you fill up the whole 6.6-meter third stage with them so they have no room for gimbal, their gravity drag losses are too high to hit TLI.
• If you swap out the RD-180s for Merlin 1Ds, you need only 36 due to their high T/W and the fact that they mean densified propellant in the first stage.
• If you bump up the third stage to the same 10.1-m diameter as the rest of the Saturn V, you can get some very nice results by using kerolox on the first stage, methalox on the second stage, and hydrolox on the third stage. With no change other than that increased third-stage diameter, you can throw 75 tonnes to TLI with 41 Merlin 1Ds, 6 Raptors, and 5 BE-3Us.
• If you imagine an expendable 3-stage Starship with the same first and second stage sizes, 8 Vac Raptors on the second stage, and clone the second stage as a third stage with 5 Vac Raptors, you could throw 100 tonnes to TLI.
• The Superheavy Booster, stripped of its grid fins and landing legs, could put 39 tonnes into LEO as an SSTO.

 

[wumpus]

22 hours ago, farmerben said:

First stage TWR was a measly 1.15

Most KSP players want TWR between 1.5 - 2.  And have flown with TWR's well above and below that, and know what it's like.    

KSP doesn't penalize Q_max(maybe RSS players know when their rocket goes through Q_max, for nearly all KSP players, it is a non-issue). While I've always believed KSP TWR should approach 2, I didn't think that was popular amoung most >1.0 players (the pre-release edition "souposphere" worked best with TWR==2.0 the whole way up).

One of the big differences between KSP and reality/NASA is that NASA (and other pro rocket builders) have fuel_mass/fueltank_mass  ratios similar or less than the ratio to a coke can and the coke inside.  TWR of 1.15 isn't a big problem (although SpaceX seems to like their TWR~=1.3 or more).  The only orbital rockets I'm aware of with TWR>2 are all-solid rockets (I think Japan has launched a few with TWR of around 5). 

[Bill Phil]

2 hours ago, sevenperforce said:

A simple Saturn V clone with the same tank sizes (but with modern construction) would throw 53 tonnes to TLI (4.4 tonnes more than Apollo 17) with 7 RD-180s on the first stage, 7 BE-3Us on the second stage, and 3 BE-3Us on the third stage.

This seems off. I may be doing things wrong, but I'm consistently getting around 70 tonnes to TLI.

I used your mass numbers for the Saturn V stages and the same specific impulses and calculated the payload to TLI. It was very close to the known value, within the margin of roundoff error (on the order of kilograms). So my analysis of the Saturn V should check out.

Then I calculated the payload using the RD-180 and BE-3U specific impulses and got nearly 70 tonnes to TLI. I didn't make any changes to stage dry or wet masses though, and I assumed the flight profile and the target delta-v (before drag) were both the same. The margin of error would be larger here since I didn't take into account engine mass. But I also didn't take into account new materials which should allow for lighter construction. Now there are other assumptions here, but around 60 to 65 tonnes to TLI seems reasonable, taking into account adapters, fairings, and so on. 

My first thought is that the choice of the BE-3U actually hurts performance almost as much as it helps, and some quick math shows it has a lower T/W than the J-2 (49 for the BE-3U and 73 for the J-2). This could make the total losses larger and would seem to hurt upper stage dry mass. You should add the HG-3 engine (or a "scaled down" SSME) and see how that performs with your analysis method for the upper stages:

https://en.wikipedia.org/wiki/HG-3_(rocket_engine)

The HG-3 has a similar specific impulse to the BE-3U but a higher T/W, and so it should help performance more than it hurts it. For the first stage the total mass is so large that a lower T/W ratio doesn't hurt performance as much, but the RD-180 does have a lower TWR than the F-1. Using the Merlin 1D should offset that but its lower specific impulse may hurt performance more. You should also add the RD-191:

https://en.wikipedia.org/wiki/RD-191

It's T/W is near the F-1's and has similar specific impulse to the RD-180.

[RCgothic]

2 hours ago, sevenperforce said:

 

I created a "sandbox" version of the spreadsheet here if people want to play around with it: https://docs.google.com/spreadsheets/d/1UNuPfjZE-YVIrmLINiUMcOzJk-voZ2TqM6R7MgyaM-I/edit?usp=sharing

If you want to change the other parameters you'll need to copy it into your own drive and change the permissions.

I think your #of engines logic may need another visit. Or at least explaining. I can fit 174 Merlin 1Ds under a 12m diameter.

38 in a 11.1m ring.

32 in a 9.3m ring.

26 in a 7.5m ring.

19 in a 5.7m ring.

13 in a 3.9m ring.

7 in a 2.1m ring

1 in the centre.

[sevenperforce]

4 minutes ago, RCgothic said:

I think your #of engines logic may need another visit. Or at least explaining. I can fit 174 Merlin 1Ds under a 12m diameter.

38 in a 11.1m ring.

32 in a 9.3m ring.

26 in a 7.5m ring.

19 in a 5.7m ring.

13 in a 3.9m ring.

7 in a 2.1m ring

1 in the centre.

My logic ran up to 20 and then defaulted to an assumed honeycomb pattern above that, but the math may be wanting. Lemme look up higher-order circle packing.

[RCgothic]

4 minutes ago, sevenperforce said:

My logic ran up to 20 and then defaulted to an assumed honeycomb pattern above that, but the math may be wanting. Lemme look up higher-order circle packing.

Mine isn't optimised packing. I just worked in giving each ring of engines a full diameter clearance, rounding down the number that would fit around that circumference.

It's definitely restricted gimballing.

[sevenperforce]

1 minute ago, RCgothic said:

Mine isn't optimised packing. I just worked in giving each ring of engines a full diameter clearance, rounding down the number that would fit around that circumference.

It's definitely restricted gimballing.

I think I went linear after 20 when I should have gone quadratic. Redoing some math.

[tater]

On 4/25/2020 at 9:26 AM, sevenperforce said:

S-II was a friggin' workhorse.

J-2... Should never have stopped making those.

[sevenperforce]

2 hours ago, RCgothic said:

Mine isn't optimised packing. I just worked in giving each ring of engines a full diameter clearance, rounding down the number that would fit around that circumference.

It's definitely restricted gimballing.

I originally had an "add six engines for each additional engine diameter" equation, which obviously produces the wrong result. I just updated it to calculate based on hexagonal area, so it will fault if the number of engines exceeds of the number of engine-width hexagons that can be inscribed in the stage cross-section. 

[sevenperforce]

3 hours ago, Bill Phil said:

This seems off. I may be doing things wrong, but I'm consistently getting around 70 tonnes to TLI.

I used your mass numbers for the Saturn V stages and the same specific impulses and calculated the payload to TLI. It was very close to the known value, within the margin of roundoff error (on the order of kilograms). So my analysis of the Saturn V should check out.

Then I calculated the payload using the RD-180 and BE-3U specific impulses and got nearly 70 tonnes to TLI. I didn't make any changes to stage dry or wet masses though, and I assumed the flight profile and the target delta-v (before drag) were both the same. The margin of error would be larger here since I didn't take into account engine mass. But I also didn't take into account new materials which should allow for lighter construction. Now there are other assumptions here, but around 60 to 65 tonnes to TLI seems reasonable, taking into account adapters, fairings, and so on. 

My first thought is that the choice of the BE-3U actually hurts performance almost as much as it helps, and some quick math shows it has a lower T/W than the J-2 (49 for the BE-3U and 73 for the J-2). This could make the total losses larger and would seem to hurt upper stage dry mass. You should add the HG-3 engine (or a "scaled down" SSME) and see how that performs with your analysis method for the upper stages:

https://en.wikipedia.org/wiki/HG-3_(rocket_engine)

The HG-3 has a similar specific impulse to the BE-3U but a higher T/W, and so it should help performance more than it hurts it. For the first stage the total mass is so large that a lower T/W ratio doesn't hurt performance as much, but the RD-180 does have a lower TWR than the F-1. Using the Merlin 1D should offset that but its lower specific impulse may hurt performance more. You should also add the RD-191:

https://en.wikipedia.org/wiki/RD-191

It's T/W is near the F-1's and has similar specific impulse to the RD-180.

I went in and added the J-2, the F-1, and the RD-191 for comparison. I added most of the stats for the HG-3 but I wasn't able to find a good source for its dry weight; if you see one, let me know. 

Any other engines you want to see?

Using the F-1 and J-2 with the other info from the Saturn V is giving me too little dV on the second stage, by a long shot. I am going to look at my structural fractions and propellant volume to see if I can figure out why.

[sevenperforce]

Ahhh, I bet I mucked up volume by taking stage lengths as inclusive of engine length and not testing against actual volume. Will rework. 

[magnemoe]

22 hours ago, wumpus said:

KSP doesn't penalize Q_max(maybe RSS players know when their rocket goes through Q_max, for nearly all KSP players, it is a non-issue). While I've always believed KSP TWR should approach 2, I didn't think that was popular amoung most >1.0 players (the pre-release edition "souposphere" worked best with TWR==2.0 the whole way up).

One of the big differences between KSP and reality/NASA is that NASA (and other pro rocket builders) have fuel_mass/fueltank_mass  ratios similar or less than the ratio to a coke can and the coke inside.  TWR of 1.15 isn't a big problem (although SpaceX seems to like their TWR~=1.3 or more).  The only orbital rockets I'm aware of with TWR>2 are all-solid rockets (I think Japan has launched a few with TWR of around 5). 

Low TWR rockets are hard to fly manual but real rockets uses an autopilot. 
Fuel and longer tanks are cheap compared to engines and yes dry mass faction is lower. 

Now if you want to reuse your first stage you want to get second stage up to it velocity faster. Starship aims for an TWR above 1.5 as they need boost back and the size of the first stage matter here. 

 

[JoeSchmuckatelli]

On 4/22/2020 at 8:26 AM, Hannu2 said:

. Of course you can build modern rocket about the same size and call it Saturn VI for nostalgic reasons, but technical solutions will be very different.

 

Do we even really need a similar size? 

 

Haven't we been using Delta 2s for Mars probes?  

 

Edited April 28, 2020 by JoeSchmuckatelli