Author Topic: MCT Speculation and Discussion Thread 4  (Read 621490 times)

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3479
  • Florida
  • Liked: 1887
  • Likes Given: 223
Re: MCT Speculation and Discussion Thread 4
« Reply #2340 on: 06/18/2016 03:24 PM »
Just wondering, how hard is methalox propellant transfer in microgravity compared to hypergolic or hydrolox refueling?

Unknown as nobody has done any.
Hydrolox would be the toughest mainly because of hydrogen's tiny atoms finding leaks really well and its need for really low temperature.
Hypergolic might be the easiest...until you accidentally started it up.

ISS is refueled by Progress.
ULA did a LOX transfer test on a Centaur several years ago. The result was that cryo transfer was simple. Just need a pressure differential between tanks and a constant ulage motor firing to keep the cryo settled. The rest is just isolation valves opening and closing and pressures of the tanks control.

Added: BTW: The key difference is that for the cryo is that there are no tank bladders but on Progress the storeable prop tanks have bladders so that no ulage motor is required.
« Last Edit: 06/18/2016 03:29 PM by oldAtlas_Eguy »

Offline docmordrid

  • Senior Member
  • *****
  • Posts: 4978
  • Michigan
  • Liked: 2099
  • Likes Given: 1
Re: MCT Speculation and Discussion Thread 4
« Reply #2341 on: 06/18/2016 03:45 PM »
Thin Red Line Aerospace has expandable cryo tank tech which sounds good for Progress style tanker/depot bladders.

Space News link...

« Last Edit: 06/18/2016 03:47 PM by docmordrid »
DM

Offline mvpel

  • Full Member
  • ****
  • Posts: 1124
  • New Hampshire
  • Liked: 1295
  • Likes Given: 1686
Re: MCT Speculation and Discussion Thread 4
« Reply #2342 on: 06/19/2016 10:30 PM »
I'm saying that I do think it would be impossible for it all to be ready for launch in six years, with no one hearing a peep of any such activity already under way.

I would love to be proven wrong, though. :)

Suffice it to say that you'd be amazed what kind of development work takes place without anyone hearing a peep of any such activity already under way. And I, from my own perspective, would love to be proven right.
« Last Edit: 06/19/2016 10:36 PM by mvpel »
"Ugly programs are like ugly suspension bridges: they're much more liable to collapse than pretty ones, because the way humans (especially engineer-humans) perceive beauty is intimately related to our ability to process and understand complexity. A language that makes it hard to write elegant code makes it hard to write good code." - Eric S. Raymond

Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2343 on: 06/23/2016 02:45 PM »
Actually, retractable nozzles usually have an expansion ratio of ~50:1, but an extendable nozzle with the retraction split pulled back to about 15% of the nozzle length gets the retracted expansion ratio down to ~20:1 so the SL ISP goes up to ~300s and the thrust to ~2000 kN if they run the chamber at 12MPa. That ISP increase is enough to reduce fuel requirements to ~650t so that a very lightweight (~42t) 4-engine vehicle could put itself in a low parking orbit for refueling.

I don't think anyone has flown a engine like that, and it has it own set of engineering difficulties. But that certainly doesn't mean SpaceX wouldn't try.
They may want an engine like that for the center landing engine in any case.

Pulling this over from the BFR launch pad thread, since it's really BFS/MCT speculation.

Some estimates at what a Raptor flying on Falcon Heavy would look like (in the FH thread) put a different light on BFS as a potential self-SSTO. The ISP numbers Tom Mueller quoted for the SL Raptor engine (321s SL, 363s vacuum) are only achievable at much higher chamber pressures than Merlin, around 20-25 MPa. At those chamber pressures a RL10-B style engine becomes considerably more feasible for SSTO operation, with a retracted average ISP of 330 to 340s. Whether rapid reusability and long life are possible at those chamber pressures is questionable, but SpaceX clearly intends to try.

Specifically at 25 MPa, 3.45 O/F, 55:1 expansion (retracted), it would get 220 tonnes thrust at a 316s SL and 359s vac ISP, increasing to 265t and 380s when the vac bell is extended. 4 of those could theoretically put a 51t ship in low parking orbit, with a liftoff mass of 700t.

The other interesting factor is that a 4 to 5 meter vac bell is unnecessary at these chamber pressures. A 25 Mpa engine can realistically get 265t of thrust at 380s ISP with only a 3.15m nozzle - only slightly larger than Mvac.

Offline Burninate

  • Full Member
  • ****
  • Posts: 1134
  • Liked: 351
  • Likes Given: 73
Re: MCT Speculation and Discussion Thread 4
« Reply #2344 on: 06/23/2016 08:18 PM »
Actually, retractable nozzles usually have an expansion ratio of ~50:1, but an extendable nozzle with the retraction split pulled back to about 15% of the nozzle length gets the retracted expansion ratio down to ~20:1 so the SL ISP goes up to ~300s and the thrust to ~2000 kN if they run the chamber at 12MPa. That ISP increase is enough to reduce fuel requirements to ~650t so that a very lightweight (~42t) 4-engine vehicle could put itself in a low parking orbit for refueling.

I don't think anyone has flown a engine like that, and it has it own set of engineering difficulties. But that certainly doesn't mean SpaceX wouldn't try.
They may want an engine like that for the center landing engine in any case.

Pulling this over from the BFR launch pad thread, since it's really BFS/MCT speculation.

Some estimates at what a Raptor flying on Falcon Heavy would look like (in the FH thread) put a different light on BFS as a potential self-SSTO. The ISP numbers Tom Mueller quoted for the SL Raptor engine (321s SL, 363s vacuum) are only achievable at much higher chamber pressures than Merlin, around 20-25 MPa. At those chamber pressures a RL10-B style engine becomes considerably more feasible for SSTO operation, with a retracted average ISP of 330 to 340s. Whether rapid reusability and long life are possible at those chamber pressures is questionable, but SpaceX clearly intends to try.

Specifically at 25 MPa, 3.45 O/F, 55:1 expansion (retracted), it would get 220 tonnes thrust at a 316s SL and 359s vac ISP, increasing to 265t and 380s when the vac bell is extended. 4 of those could theoretically put a 51t ship in low parking orbit, with a liftoff mass of 700t.

The other interesting factor is that a 4 to 5 meter vac bell is unnecessary at these chamber pressures. A 25 Mpa engine can realistically get 265t of thrust at 380s ISP with only a 3.15m nozzle - only slightly larger than Mvac.

On that note, pulling this over from the FH thread, because Raptor speculation has direct import on MCT's mission architecture:

That's interesting. So the best way to increase the payload of the Falcon Heavy is a bigger upper stage. If such a need existed, I would guess space X would choose to stretch the second stage, and add two more vaccum Merlins to the stage. Easier than integating a second fuel type into the launch infrastructure to allow for the use of the Raptor engine.
No need to add an engine, the Merlin has plenty of thrust for a larger upper stage, especially if that stage is only used with the FH so it will be further along than the current F9 upper stage is at ignition.

The reason to go Raptor is to get improved ISP and more balanced fuel/oxidizer temperatures (which may make long life easier without adding extra mass).

Plus the Merlin Vacuum engine is gigantic, there is no way to fit 3 of them on any remotely Falcon-sized stage. You'd need a 8m diameter interstage at minimum.

To re-iterate that point.



MVac pushes 95 tonnes. By the time the fuel runs out, the engine has to throttle below 40% so as to not kill the payload with high g-forces.
The feasible way is to have a wider upper stage and interstage.
This probably is what they will do if they go wider. But I think it will take some careful thought, since the widest part of the nozzle is at the bottom. And the bottom is where the narrowest part of the interstage would be, since that is where it starts flaring out... so there may need to be a gap of some size to get it to work.
...or have an extendible nozzle like the part that you trimmed from my post.

The last bit of the conversation is great.

Extendible nozzle helps with interstage length, doesn't nothing for width. 

A 4.8 meter nozzle is 4.8 meters whether in 1, 2 or more pieces.  As we've seen from on board footage the separation events are not that smooth.

A rocket with a 500,000 lbf upper stage is going to need good clearances and some creative problem solving.

Of course maybe there is a smaller Raptor in the works and this discussion changes.

How much ISP would Raptor lose when it has a nozzle diameter of only 4m?
Does it have to make the same thrust as the 4.8m nozzle? Because ISP isn't solely dictated by nozzle size - for the same fuels it is a function of pressure ratios... and there are other ways to change pressure ratios (e.g. change mass flow and/or throat diameter) at the expense of thrust.

Obviously I don't even know enough to ask the right question. For arguments sake assume the same engine, just a smaller nozzle.

Though what we know does not rule out that a dedicated smaller version of Raptor might be built.
Raptor with a 4 m nozzle loses about 1% of ISP compared to a 4.8 m nozzle: 376 s vs 380 s.

This is based on sim in RPA lite using: Methane/LOX at:
9.7 MPa chamber pressure (same as Merlin)
2.8 O/F ratio (optimum for methalox at 9.7 MPa)
165 expansion ratio for the 4.8 m nozzle (same as Merlin Vac)
115 expansion ratio for the 4 m nozzle (assuming same throat diameter as the 4.8 m nozzle)

O/F ratio was supposed to be 3.8.  Chamber pressure seems like the big unknown - if the M1D gas generator produces 9.7MPa, what should we expect in a FFSC Raptor?

Musk: "The critical elements of the solution are rocket reusability and low cost propellant (CH4 and O2 at an O/F ratio of ~ 3.8 ). And, of course, making the return propellant on Mars, which has a handy CO2 atmosphere and lots of H2O frozen in the soil." - http://waitbutwhy.com/2015/08/how-and-why-spacex-will-colonize-mars.html/4


Neither of those particularly affect the relative differences between the 4 meter and 4.8 meter nozzle; the benefit of having extra expansion is about the same for a higher pressure engine.

I just grabbed that 2.8:1 O/F number off a chart, and according to RPA it is actually a bit low for a 165:1 expansion vacuum engine. At 10 MPa the O/F for optimal ISP is 3.4:1, and at 15 MPa it optimizes at 3.45:1. However, 3.8:1 is rather high and actually loses a second or two of ISP (although you get a bit better thrust due to higher massflow).


How are real numbers likely to come in, in relation to the figures provided by RPA Lite?  Would the RPA numbers represent a theoretical maximum or perhaps a plausible best-guess at the exact values?

RPA calculates both the theoretical best possible performance, and a best guess at the actual performance after accounting for combustion and nozzle inefficiencies. I have no idea how accurate it is in reality. And we don't have a ton of known parameters to put in.
From the O/F and the two isp points I had estimated 20.5MPa as Pc. Can't recall the expansion.
The four datapoints we have are 321s (sealevel small-nozzle?) , 363s (vacuum small-nozzle?), and 380s (vacuum big-nozzle?), with an O:F ratio of 3.8.  I had 14.5MPa working for theoretical numbers ("Express thermodynamic analysis") at expansion ratios of 32.5 and 75.

Changing it to "Extended analysis" and examining chamber performance given the estimated reaction and nozzle efficiencies, I end up losing 14s off of each number.
I used actual vs theoretical. And modified a bit the freeze chemistry since the ch4/lox combustion, while appears simple, has a lot of steps.
Punch it up to the RD-180 chamber pressure - 26.7MPa
Using the listed estimates for 3.8:1 inefficiencies rather than theoretical maxima in RPA Lite:
With an expansion ratio of 59, you get 321s at sealevel and 363s in vacuum.
With an expansion ratio of 164, you get 380s in vacuum.

Sidenote: Do we have a higher chamber pressure liquid rocket engine in existence than the RD-180?

That's very interesting. Do you really think they will try to run at 26.7 MPa? For comparison, the RD-191 and RD-180 run at 26.7 MPa, and the SSME ran up to 21 MPa. I think that's the highest pressure ever flown in a reusable engine. A 100% length bell nozzle gets the same performance at 25 MPa and 55:1 expansion (155:1 for the vac nozzle).

The nice thing about higher chamber pressures is the engine gets much smaller (but not lighter) for the same thrust. At 25 MPa the SL Raptor only needs a 1.88m diameter nozzle to get Musk's estimated 500klbf of thrust, and the vac version gets 591klbf with only a 3.16m dia nozzle.

To bring this entire conversation back to Falcon Heavy (we were wandering OT for a bit there), that would mean it's possible a 25 MPa Raptor Vac could fly in the current Falcon interstage. Even with the nozzle trimmed back to 70% (which would put it very nearly in the Mvac envelope), it could realistically get 265 tonnes of thrust at 377s ISP. That's a healthy upgrade over Mvac's 95 tonnes thrust and 348 ISP.

I have no idea.  I was about to ask you, and everybody else.

What chamber pressures should we expect out of Raptor?  Do we have any evidence or reasoning one way or the other? (other than the previously mentioned 321s, 363s, 380s, and 3.8:1, and "FFSC should permit higher chamber pressures" notion)
Sidenote: Do we have a higher chamber pressure liquid rocket engine in existence than the RD-180?
RD-191, 262.6kg/cm² vs 261.7kg/cm². RD-270 was 266.1.  I know of nothing bigger than this. But some military RCS engines (like the one in MIRVs) probably have higher Pc.

Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2345 on: 06/23/2016 08:43 PM »
Thanks for collecting those. As far as the expected chamber pressure, that really depends on the performance and design requirements. Increasing the pressure requires a bigger or faster turbopump... and a lost more strength to keep everything from going kabloomy.

However, the 321s SL and 363s vacuum ISP numbers do put some hard limits on how low chamber pressure can be. Getting 363s in vacuum requires at least 50:1 expansion even with a very efficient nozzle design... but getting 321s with 50:1 expansion agasint SL backpressure requires at least 22.5 MPa chamber pressure. So a pressure in the 22-25 MPa range looks pretty likely based on that info.

Offline Impaler

  • Full Member
  • ****
  • Posts: 1283
  • South Hill, Virgina
  • Liked: 363
  • Likes Given: 0
Re: MCT Speculation and Discussion Thread 4
« Reply #2346 on: 06/23/2016 10:14 PM »
I had always though the estimates of nozzle diameter for the Raptor were WAY too high, if were looking at vac nozzles only around 3 m across then a hexagonal 6 engine arrangement can easily fit in an upper-stage of around 12.5 m in diameter.

Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2347 on: 06/24/2016 01:49 AM »
We really need a Raptor speculation thread  ::)

If Raptor follows a development arc somewhat like Merlin, the first versions will probably operate at lower pressure/thrust/impulse than later versions, as SpaceX figures out methane, staged combustion, high pressure engines, etc. I could see a MVac-sized version flying on Falcon Heavy with a stretched or bigger upper stage, maybe on Falcon 9 as well.

BFS won't have the packaging constraints or the need to throttle as deep to avoid excessive acceleration.

Offline Andy USA

  • Lead Moderator
  • Administrator
  • Full Member
  • *****
  • Posts: 981
  • Los Angeles, California
  • Liked: 129
  • Likes Given: 126
Re: MCT Speculation and Discussion Thread 4
« Reply #2348 on: 06/25/2016 03:24 AM »
Thread trimmed as it was taken off topic by a completely false statement that was removed from another thread and then reposted here. I will be reviewing the members in question.

Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2349 on: 06/27/2016 06:42 PM »
I had always though the estimates of nozzle diameter for the Raptor were WAY too high, if were looking at vac nozzles only around 3 m across then a hexagonal 6 engine arrangement can easily fit in an upper-stage of around 12.5 m in diameter.

You don't think 6 engines is a bit much, even for engine-out operation? It looks like Raptor vac should make 265 tonnes thrust each, if the SL version makes 230. That would be a total vacuum thrust of 1600 tonnes with 6 engines. If it's designed to TWR=1 at 1.5 kms staging with a 92% PMF, that puts 300t total in LEO (and fully refueled in about eight 175t tanker trips would put 300t total through fast TMI and to the Mars surface).

A 4-engine design with 92% PMF and TWR=1 at 1.5 kms staging would lift 100t of payload to a 500km parking orbit with 5% margins, even with 2 engines out. It would still be capable of pulling a 3g Mars landing burn even with 2 engines out (530t thrust, 180t mass at burnout), and more than 4g at Earth landing even accounting for SL backpressure. And it would have a 1.47:1 TWR at Mars liftoff with a full 770t prop load.

While it would be nice to get to LEO with 100t payload and 50-80t of prop left, I just don't see that being worthwhile if it means the spacecraft is grossly oversized for everything else.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28475
  • Minnesota
  • Liked: 8347
  • Likes Given: 5482
Re: MCT Speculation and Discussion Thread 4
« Reply #2350 on: 06/27/2016 06:49 PM »
I doubt that more than 5 Raptors are required.
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 biosehnsucht

  • Full Member
  • **
  • Posts: 292
  • Liked: 101
  • Likes Given: 169
Re: MCT Speculation and Discussion Thread 4
« Reply #2351 on: 06/27/2016 10:02 PM »
Wouldn't depending on which engines are 'out', fewer number of engines could potentially require greater gimball and/or throttle ranges to compensate - i.e. if you have 4 engines and capable of 2-out but the 2 you lose are next to each other, you're gonna want some Shuttle-esque gimbal? With 6 engines, you can probably get away with lower gimbal ranges, etc.

Offline Impaler

  • Full Member
  • ****
  • Posts: 1283
  • South Hill, Virgina
  • Liked: 363
  • Likes Given: 0
Re: MCT Speculation and Discussion Thread 4
« Reply #2352 on: 06/27/2016 11:21 PM »
My main point was that you can get all the thrust you could want in a 12.5 m diameter upper stage and engine thrust density is not a driver for any higher diameter.

Having an engine out capability on a 2nd stage would be a first and very desirable as the intent is to reuse them.  I am very doubtful that gimbaling on remaining engines will be adequate to compensate for any engine out because the vehicle is too short, that puts the center of mass closer to the engines and the thrust vector MUST go through the center of mass of the vehicle, that is MUCH easier on a long thin first stage.  Thus I conclude that a symmetrical shutdown is the only viable means to get engine out, the remaining 4 engines of a 6 engine configuration would be sufficient to reach orbit.


Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2353 on: 06/28/2016 03:21 AM »
I was implying symmetrical shutdown in the two engine out scenario. With 2 of 4 engines out as described it still has plenty of margin to orbit. It would be pretty unlikely to lose opposite pairs simultaneously, IMO. I think the numbers still work out better for the 4 engine case, unless Raptor vac turns out to be a lot smaller than expected.

Offline Impaler

  • Full Member
  • ****
  • Posts: 1283
  • South Hill, Virgina
  • Liked: 363
  • Likes Given: 0
Re: MCT Speculation and Discussion Thread 4
« Reply #2354 on: 06/28/2016 08:24 AM »
Your payload is too small at 100 mt, I'm thinking payload to LEO of around 200 mt even after engine loss, so that would require 4 functional engines and 6 total.

Offline envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2355 on: 06/28/2016 02:00 PM »
Your payload is too small at 100 mt, I'm thinking payload to LEO of around 200 mt even after engine loss, so that would require 4 functional engines and 6 total.

But why would you want that? Staging to LEO and LEO to Mars surface have almost exactly the same DV requirements, so the vehicle should be optimized to deliver the same payload to both. A typical fast transit and EDL only requires ~6 kms. Launching 200t payload and 100t ship into LEO requires about 1400t of propellant in the US, but transit and EDL of 100t payload and 100t ship only require about 800t. You're shipping an extra 20% to 30% of useless dry mass (engines and tankage) to Mars and back.

For LEO tanker runs the 200t payload is useful, but that can be accomplished with larger tanks that the Mars transit ship doesn't need. And it still doesn't really need 6 engines for high-thrust engine out capability, because a tanker won't be carrying people and it can always use some of it's extra fuel load for margin. An off-nominal LEO launch has a lot of options for abort, particularly if a rapid launch cadence, on-orbit refueling, and LEO rendezvous are SOP.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28475
  • Minnesota
  • Liked: 8347
  • Likes Given: 5482
Re: MCT Speculation and Discussion Thread 4
« Reply #2356 on: 06/28/2016 02:42 PM »
Yeah, I'm fairly certain that the BFS will go straight to Earth from the surface of Mars (at least for the early missions), in which case you'll want the BFS fully fueled with 1000-1400 tons of propellant.
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 envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2357 on: 06/28/2016 04:26 PM »
How much dV are you estimating for return, and how fast a transit? A typical 6-month return is only about 6.5 to 7 km/s from Mars surface to Earth surface (5.25 for launch to escape, 1 to 1.5 for transfer injection, 0.35 for EDL).

Increasing return DV reduces the total mission duration by allowing a earlier launch, but I don't see any solutions in the trajectory browser for reducing the actual return transit time below about 5 months. If you have computed such solutions I'd love to see them.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28475
  • Minnesota
  • Liked: 8347
  • Likes Given: 5482
Re: MCT Speculation and Discussion Thread 4
« Reply #2358 on: 06/28/2016 05:10 PM »
How much dV are you estimating for return, and how fast a transit? A typical 6-month return is only about 6.5 to 7 km/s from Mars surface to Earth surface (5.25 for launch to escape, 1 to 1.5 for transfer injection, 0.35 for EDL).

Increasing return DV reduces the total mission duration by allowing a earlier launch, but I don't see any solutions in the trajectory browser for reducing the actual return transit time below about 5 months. If you have computed such solutions I'd love to see them.
6.5-7km/s sounds about right. I expect return trips to be much longer than 100 days.
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 envy887

  • Senior Member
  • *****
  • Posts: 4622
  • Liked: 2539
  • Likes Given: 1395
Re: MCT Speculation and Discussion Thread 4
« Reply #2359 on: 06/28/2016 06:05 PM »
How much dV are you estimating for return, and how fast a transit? A typical 6-month return is only about 6.5 to 7 km/s from Mars surface to Earth surface (5.25 for launch to escape, 1 to 1.5 for transfer injection, 0.35 for EDL).

Increasing return DV reduces the total mission duration by allowing a earlier launch, but I don't see any solutions in the trajectory browser for reducing the actual return transit time below about 5 months. If you have computed such solutions I'd love to see them.
6.5-7km/s sounds about right. I expect return trips to be much longer than 100 days.

Then your dry mass or return payload is too high if it needs 1000-1400t of prop for direct return. A 80t dry, 380s ISP vehicle with 25t return payload gets 7 km/s with less than 600t of prop, and can do 6.5 km/s with only 500t. Even a 100t vehicle with a 50t return payload only needs 725t to 825t of prop.

I can't see a real need to ever have more than ~800t of methalox in a crewed version of the BFS, as long as the dry mass is reasonable. That does all three legs of the trip (BFR staging to LEO; LEO to Mars surface; Mars surface to Earth surface) with a lot more than the expected payload for each leg.

Tags: