Author Topic: SpaceX: Merlin 1D Updates and Discussion Thread 2  (Read 248832 times)

Offline Prober

  • Senior Member
  • *****
  • Posts: 10316
  • Save the spin....I'm keeping you honest!
  • Nevada
  • Liked: 700
  • Likes Given: 728
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #40 on: 12/14/2013 01:15 PM »


I'm working on a 3D printable model of the F9 v1.1...

Great news and looking forward to printing your model in the future  :)
2017 - Everything Old is New Again.
I have not failed. I've just found 10,000 ways that won't work. ~ by Thomas Alva Edison

Offline pagheca

  • Bayesian Pundit. Maybe.
  • Full Member
  • ****
  • Posts: 759
  • Lives in Ivory, Tower
  • Liked: 213
  • Likes Given: 161
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #41 on: 12/14/2013 01:57 PM »
Hmm, SpaceX Falcon9 main page reports this:
"The first stage engines are gradually throttled near the end of first-stage flight to limit launch vehicle acceleration as the rocketís mass decreases with the burning of fuel."
http://www.spacex.com/falcon9

Ops... I someway missed this information. Available data from first v1.1 flight on altitude and velocity fit quite well with a full throttle simulation with a reasonable trajectory. However, I noticed that near the end data completely miss the model. Thanks to your hint I searched the internet and found that "After passing the T+2-minute mark, the Falcon 9 started to throttle back its engines to limit stress on the vehicle" (from http://www.spaceflight101.com/falcon-9-v11-cassiope-launch-updates.html). If this information is accurate, the fit for the first 2 mins can be used, by including a careful estimation of the gravity and aerodynamic drag, to evaluate the trajectory of the rocket, that is actually what I'm trying to do now.

thanks again

p.s. Yes, I imagine this has been already done by many others.
« Last Edit: 12/14/2013 01:59 PM by pagheca »

Online ugordan

  • Senior Member
  • *****
  • Posts: 7259
    • My mainly Cassini image gallery
  • Liked: 1400
  • Likes Given: 289
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #42 on: 12/14/2013 02:03 PM »
Available data from first v1.1 flight on altitude and velocity fit quite well with a full throttle simulation with a reasonable trajectory.

Because MECO-1 happened earlier than on an expendable flight, roughly at the time it would normally start throttling down due to G limits. I would expect the SES-8 flight did involve throttling to some degree.

Offline pagheca

  • Bayesian Pundit. Maybe.
  • Full Member
  • ****
  • Posts: 759
  • Lives in Ivory, Tower
  • Liked: 213
  • Likes Given: 161
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #43 on: 12/14/2013 02:15 PM »
Available data from first v1.1 flight on altitude and velocity fit quite well with a full throttle simulation with a reasonable trajectory.

Because MECO-1 happened earlier than on an expendable flight, roughly at the time it would normally start throttling down due to G limits. I would expect the SES-8 flight did involve throttling to some degree.

I'm not completely sure I understand what you mean. However, MECO happened at T+163 on this flight. 43 sec after throttling down started if the 2 min marker is taken for granted. However, a fit with the first 2 mins is enough to get some indication on the trajectory to be checked with available imaging.

Online ugordan

  • Senior Member
  • *****
  • Posts: 7259
    • My mainly Cassini image gallery
  • Liked: 1400
  • Likes Given: 289
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #44 on: 12/14/2013 02:21 PM »
43 sec after throttling down started if the 2 min marker is taken for granted. However, a fit with the first 2 mins is enough to get some indication on the trajectory to be checked with available imaging.

Simple. The "after passing the 2 minute mark" is vague, it doesn't imply throttling immediately after 2 min nor would that be consistent with vehicle acceleration at that time.

Take v1.0 as an example, with a roughly similar split between stages. It shut down 2 engines some 15 seconds before MECO-2 so around 165 seconds or thereabouts. 15 - 20 seconds before running dry is roughly the time Cassiope staged.

Offline fatjohn1408

  • Full Member
  • ***
  • Posts: 313
  • Liked: 13
  • Likes Given: 13
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #45 on: 12/20/2013 10:04 AM »
Hi everyone,

this question maybe obvious for many of you - please excuse me if this is the case but I'm just in learning mode.

As well known, the Merlin 1D engine can be throttled down from 100% to 60%.

The vacuum thrust of a rocket engine is proportional to the exhaust velocity and to the mass flow rate of the gas exhaust: dm/dt * V_exh.

In vacuum and if also the gravity drag is negligible, when a Merlin 1D engine is throttled down, is dm/dt only changing, OR also V_exh (aka Isp_vacuum) is affected someway, with a consequent reduction in the engine efficiency in using the available propellant?

If the second hypothesis is right (as I would say by looking at the physical principles the engine is based on), does someone know a reasonable way to write the relation Thrust(%) = f(%,dm/dt, V_exh)?

Thanks for any hint,

Changing thrust settings will change the chamber pressure, the chamber temperature, the characteristic velocity and correspondingly the specific impulse. And in the process also the mass flow.
It will leave key characteristics unchanged, such as the exit area, throat area, the oxidizer-to-fuel ratio and thus the ratio of specific heats , the vandenkerkhove function, the exit to chamber pressure ratio and therefore the thrust coefficient.

Since the throat area and the thrust coefficient stay constant, a change in thrust leads to a linearly corresponding change in chamber pressure and from this pressure ratio, a change in chamber temperature can be found using Poissonís equation

This leads in turn to a change in characteristic velocity and specific impulse.
At 50% throttle a decrease of specific impulse from 340s to 317s is to be expected

On top you can expect some losses due to combustion efficiency that gets lower with lower throttle in a pintle engine.
« Last Edit: 12/20/2013 10:06 AM by fatjohn1408 »

Offline pagheca

  • Bayesian Pundit. Maybe.
  • Full Member
  • ****
  • Posts: 759
  • Lives in Ivory, Tower
  • Liked: 213
  • Likes Given: 161
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #46 on: 12/20/2013 12:15 PM »
[Changing thrust settings will change the chamber pressure, the chamber temperature, the characteristic velocity and correspondingly the specific impulse. And in the process also the mass flow.
It will leave key characteristics unchanged, such as the exit area, throat area, the oxidizer-to-fuel ratio and thus the ratio of specific heats , the vandenkerkhove function, the exit to chamber pressure ratio and therefore the thrust coefficient.

Since the throat area and the thrust coefficient stay constant, a change in thrust leads to a linearly corresponding change in chamber pressure and from this pressure ratio, a change in chamber temperature can be found using Poissonís equation

This leads in turn to a change in characteristic velocity and specific impulse.
At 50% throttle a decrease of specific impulse from 340s to 317s is to be expected

On top you can expect some losses due to combustion efficiency that gets lower with lower throttle in a pintle engine.

Thanks for this comment.

This is more or less what I guessed. In particular I was tempted to apply the Poisson's equation. This would mean that throttling the system 50% makes it much less performant. And that 2 engines working at 50% are less savvy than a single one working at 100% in terms of fuel usage.

The problem is that I couldn't find any confirmation this is true. I only found a very old paper from 1955, telling a very different story.

Tomazic, William A. Rocket-engine throttling. UNT Digital Library. http://digital.library.unt.edu/ark:/67531/metadc61716/.

The paper is based on actual measurements performed on two different types of injection devices. I haven't followed  all the math yet, although is quite simple, but apparently, for a wide range of variation in the thrust, experimental data show that Vexh remain substantially constant.  Of course the paper is so old it doesn't mean very much, but I couldn't find anything better and the principles used today are the same.

More important, this fits with the impression that SpaceX mission profilers are not so concerned about using rockets at low throttle. For example: it's well known that during the final phase of the ascent, when the rocket has already used most of his fuel, a Falcon 9 v1.1 1st stage reduces the throttle rather than switching off more and more of his 9 Merlins 1D.

Also Space Shuttle and other LV's rockets are throttle down in some situation, when a more efficient solution would be to just switch them off. But the important point of the Merlin is that they are designed to allow a large number of restart, and that throttling happens at the end of the ascent, not - for example - during MaxQ, so that you would need to switch it on again later, that would be quite demanding in terms of reliability.

Why that? Maybe just because progressively throttling is more efficient, at the end of the day, because it allows to follow much more carefully the thrust demand than just switching off engines?

Cheers,

« Last Edit: 12/20/2013 12:23 PM by pagheca »

Offline deltaV

  • Full Member
  • ****
  • Posts: 1535
  • Change in velocity
  • Liked: 163
  • Likes Given: 453
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #47 on: 12/20/2013 04:21 PM »
I ran RPA lite with the following settings: 1000 psi chamber pressure, RP-1 and LOX with O/F=2.555, area ratio 50 with frozen equilibrium flow at area ratio 2. It gives theoretical vacuum ISP of 352.9 seconds at full throttle and 350.4 seconds at 50%, i.e., the vacuum ISP is more or less independent of throttle setting. Throttling does have a major effect on sea level performance (as expected), with 256.1 s at 50% and 278.9 s at 100%.

I believe Poisson's equation for adiabatic expansion is inappropriate here because that equation deals with adiabatic expansion that is also isentropic, which requires that work is extracted as the gas expands. If you think of the throttled engine chamber conditions as the original chamber conditions followed by a transformation, I think the appropriate transformation is an adiabatic but non-isentropic throttle-type pressure drop as seen in the Joule-Thomson effect. For ideal gases this leaves the temperature unchanged and for real gases the temperature change is small.

I seem to recall reading that the reason that higher chamber pressure slightly improves ISP is that higher chamber pressure pushes the equilibrium towards higher molecular weight products like H20 and CO2 and away from lower molecular weight ones like CO, OH-, and H+. In other words higher chamber pressure produces more complete combustion.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 25508
  • Minnesota
  • Liked: 5667
  • Likes Given: 4251
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #48 on: 12/20/2013 05:26 PM »
Well that may be true, The real reason that higher chamber pressure increases exhaust velocity is because you can get a higher expansion ratio.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline solartear

  • Member
  • Posts: 96
  • Liked: 11
  • Likes Given: 5
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #49 on: 12/20/2013 06:27 PM »
Why that? Maybe just because progressively throttling is more efficient, at the end of the day, because it allows to follow much more carefully the thrust demand than just switching off engines?

IMO a big reason would be if another engine unexpectedly shutdown, as has happened with STS and Falcon9, then they just throttle up the remaining engines, with much shorter delay and avoiding the effects of restarting an engine in flight.

Offline pagheca

  • Bayesian Pundit. Maybe.
  • Full Member
  • ****
  • Posts: 759
  • Lives in Ivory, Tower
  • Liked: 213
  • Likes Given: 161
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #50 on: 12/20/2013 09:38 PM »
IMO a big reason would be if another engine unexpectedly shutdown, as has happened with STS and Falcon9, then they just throttle up the remaining engines, with much shorter delay and avoiding the effects of restarting an engine in flight.

Yep. Good point... You may be right IMHO.
« Last Edit: 12/20/2013 09:39 PM by pagheca »

Offline Hobbes-22

  • Full Member
  • ***
  • Posts: 325
  • Acme Engineering
    • Acme Engineering shop
  • Liked: 91
  • Likes Given: 67
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #51 on: 12/27/2013 01:26 PM »
While trying to understand Merlin 1D and in particular "Merlin 1D+"* in depth, I've iterated my calculations a few times and have reached internal coherence and good balance with reality using the following characteristics/specs. 

Merlin 1D..Merlin 1D Vac..Merlin 1D+..Merlin 1D+ Vac
Nozzle diameter, m1.073.031.073.03


I've tried to draw the Falcon 9 first stage with your diameter, but I can't get it to fit. A circle of 8 engines with a diameter of 1,07 m each is going to have an outer diameter of at least 3,8 m as drawn in my CAD program, and that's with the engine nozzles touching each other. But if you look at images of the launch, the engines do not protrude outside the first stage diameter. And there's a gap between the engines.

If I limit the outer diameter of the 8 engines to 3,66 m and allow some spacing between them, the nozzle diameter is around 96,5 cm (my drawing was in 1:144).

I've also tried measuring the diameter from the second photo. Ignoring the distortion, the space between the center engine and the outer engines is 0.452 times the diameter of the center engine. So the total diameter of the ring of 8 engines is (3 + (2*0.452)) times the diameter of one engine nozzle. If the total ring diameter is 366 cm, then one engine must be 93,5 cm in diameter.

Offline malu5531

  • Full Member
  • **
  • Posts: 284
  • Sweden
  • Liked: 151
  • Likes Given: 172
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #52 on: 12/31/2013 01:39 AM »
I fully agree, nozzle diameter should be something like ~93-94 cm. I'm not fully there with the model. I have a similar issue with the RD-0162, not sure yet if those are related or there is a more trivial problem with the 1D model (such as adjusting the chamber pressure a bit, since that info might be old). 

While trying to understand Merlin 1D and in particular "Merlin 1D+"* in depth, I've iterated my calculations a few times and have reached internal coherence and good balance with reality using the following characteristics/specs. 

Merlin 1D..Merlin 1D Vac..Merlin 1D+..Merlin 1D+ Vac
Nozzle diameter, m1.073.031.073.03


I've tried to draw the Falcon 9 first stage with your diameter, but I can't get it to fit. A circle of 8 engines with a diameter of 1,07 m each is going to have an outer diameter of at least 3,8 m as drawn in my CAD program, and that's with the engine nozzles touching each other. But if you look at images of the launch, the engines do not protrude outside the first stage diameter. And there's a gap between the engines.

If I limit the outer diameter of the 8 engines to 3,66 m and allow some spacing between them, the nozzle diameter is around 96,5 cm (my drawing was in 1:144).

I've also tried measuring the diameter from the second photo. Ignoring the distortion, the space between the center engine and the outer engines is 0.452 times the diameter of the center engine. So the total diameter of the ring of 8 engines is (3 + (2*0.452)) times the diameter of one engine nozzle. If the total ring diameter is 366 cm, then one engine must be 93,5 cm in diameter.

Offline Ben the Space Brit

  • Senior Member
  • *****
  • Posts: 6755
  • A spaceflight fan
  • London, UK
  • Liked: 426
  • Likes Given: 468
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #53 on: 01/06/2014 10:26 PM »
I might just not be looking in the right place, but has the Isp of Merlin-1D-VAC been confirmed or even reliably estimated yet?
"Oops! I left the silly thing in reverse!" - Duck Dodgers

~*~*~*~

The Space Shuttle Program - 1981-2011

The time for words has passed; The time has come to put up or shut up!
DON'T PROPAGANDISE, FLY!!!

Offline Jason1701

  • Senior Member
  • *****
  • Posts: 2238
  • Liked: 70
  • Likes Given: 39
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #54 on: 01/06/2014 10:28 PM »
I might just not be looking in the right place, but has the Isp of Merlin-1D-VAC been confirmed or even reliably estimated yet?

340 s in vacuum according to their website.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 4859
  • Liked: 625
  • Likes Given: 433
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #55 on: 02/19/2014 11:16 AM »
I fully agree, nozzle diameter should be something like ~93-94 cm. I'm not fully there with the model. I have a similar issue with the RD-0162, not sure yet if those are related or there is a more trivial problem with the 1D model (such as adjusting the chamber pressure a bit, since that info might be old). 
I fully agree, nozzle diameter should be something like ~93-94 cm. I'm not fully there with the model. I have a similar issue with the RD-0162, not sure yet if those are related or there is a more trivial problem with the 1D model (such as adjusting the chamber pressure a bit, since that info might be old).

Chamber pressure need not enter into it In the 1D model*.  Thrust is

    F = q ve + Ae (pe - pa) ,

where pa is the ambient pressure.  Therefore the difference between sea-level (pa = pSL = 1 atm) thrust and vacuum (pa = 0) thrust is

    Fvac - FSL = Ae pSL ,

so

    Ae = (Fvac - FSL) / pSL .

If we take the total thrust of the Falcon 9's first stage at sea level and in vacuo from the Falcon 9 web page and divide by nine, we get single-engine thrusts of 653.9 and 741.3 kN, respectively.  For pSL = 101.325 kPa, we get Ae = 0.863 m2 and hence a diameter of 1.048 m, which agrees closely with the value you've calculated.  On the other hand, if we look at the web page for the Merlin engine itself, we're told that the engine's vacuum thrust is just 716 kN.  This lower thrust gives an exit area of 0.613 m2 and a diameter of just 0.883 m.  This fits within the geometric limit found by Hobbes-22.

Now, it could be that this simple analysis violates some constraints imposed by your more extensive model.  I would think, though (and please correct me if I'm wrong), that the the other constraints have are pretty loose, given SpaceX's reluctance to give engineering specifics.



* Actually, I suppose that's not strictly true.  If we're going to assume that flow separation occurs once the pressure drops more than a certain amount below ambient, then the effective nozzle area would depend on chamber pressure.  Thus far, though, we've been assuming there's no flow separation.  As far as I know (which isn't very far), flow separation is usually avoided these days.  (The sustainer of the classic Atlas was over-expanded at sea level to the point that flow separation did occur, but that was back when men were men :) .)  Anyway, allowing for flow separation would tend to increase our estimate of the nozzle's size.  Since we're pretty close to the size allowed by geometry already, this suggests that separation does not occur.
« Last Edit: 02/19/2014 04:20 PM by Proponent »

Offline aero

  • Senior Member
  • *****
  • Posts: 2696
  • 92111
  • Liked: 679
  • Likes Given: 214
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #56 on: 02/19/2014 09:49 PM »
There is a picture up the nozzles of the Merlin C posted here.

http://forum.nasaspaceflight.com/index.php?topic=33598.msg1161838#msg1161838

It looks to me as though SpaceX has done something to suppress flow separation at sea level. I would call it a duel bell nozzle design except it looks like quadruple bell design would be more accurate.

I can see that the nozzles are uniformly divergent on the outside but on the inside it is different. What do you all think?
Retired, working interesting problems

Offline Lars_J

  • Senior Member
  • *****
  • Posts: 6161
  • California
  • Liked: 664
  • Likes Given: 195
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #57 on: 02/19/2014 09:52 PM »
There is a picture up the nozzles of the Merlin C posted here.

http://forum.nasaspaceflight.com/index.php?topic=33598.msg1161838#msg1161838

It looks to me as though SpaceX has done something to suppress flow separation at sea level. I would call it a duel bell nozzle design except it looks like quadruple bell design would be more accurate.

I can see that the nozzles are uniformly divergent on the outside but on the inside it is different. What do you all think?

What are you drawing *that* conclusion from? That is just an engineering drawing with different sub-component coloured differently. I think you are reading more into those colours that you should.

Offline aero

  • Senior Member
  • *****
  • Posts: 2696
  • 92111
  • Liked: 679
  • Likes Given: 214
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #58 on: 02/19/2014 09:56 PM »
There is a picture up the nozzles of the Merlin C posted here.

http://forum.nasaspaceflight.com/index.php?topic=33598.msg1161838#msg1161838

It looks to me as though SpaceX has done something to suppress flow separation at sea level. I would call it a duel bell nozzle design except it looks like quadruple bell design would be more accurate.

I can see that the nozzles are uniformly divergent on the outside but on the inside it is different. What do you all think?

What are you drawing *that* conclusion from? That is just an engineering drawing with different sub-component coloured differently. I think you are reading more into those colours that you should.

You're right. That's not a picture, my bad. Does anyone have a picture of the inside of a SpaceX rocket engine nozzle?
Retired, working interesting problems

Offline Lars_J

  • Senior Member
  • *****
  • Posts: 6161
  • California
  • Liked: 664
  • Likes Given: 195
Re: SpaceX: Merlin 1D Updates and Discussion Thread 2
« Reply #59 on: 02/19/2014 10:01 PM »
There is a picture up the nozzles of the Merlin C posted here.

http://forum.nasaspaceflight.com/index.php?topic=33598.msg1161838#msg1161838

It looks to me as though SpaceX has done something to suppress flow separation at sea level. I would call it a duel bell nozzle design except it looks like quadruple bell design would be more accurate.

I can see that the nozzles are uniformly divergent on the outside but on the inside it is different. What do you all think?

What are you drawing *that* conclusion from? That is just an engineering drawing with different sub-component coloured differently. I think you are reading more into those colours that you should.

You're right. That's not a picture, my bad. Does anyone have a picture of the inside of a SpaceX rocket engine nozzle?

Here is one, with M1C's:

EDIT: added a M1D picture:
« Last Edit: 02/19/2014 10:02 PM by Lars_J »

Tags: