SpaceX: “Close, but no cigar. This time” • Hacker News Board

Re: SpaceX: “Close, but no cigar. This time”

#231

Post by jacquesm » Sat, Jan 17, 2015, 2:00 AM UTC

sillysaurus3 wrote: post #230

jacquesm wrote: post #229

Earlier quoted context omitted.

> Are you sure a few-centimeter speedbump isn't a big deal? Depends on lots of things, such as the angle of incidence of the pod relative to the direction of the wall of the tube. > I don't know how to calculate whether or not it would be, but I'm interested in knowing why you feel it won't cause problems. I don't know how to calculate rockets landing on barges either, but SpaceX seems to be able to pull it off. Ther…

That solution is interesting, but does it account for the fact that the tube curves all throughout its route? Some parts need to be curvy, some need to be straight. Are you sure that expansion can be coaxed into a curve shape rather than linear? How? Also, curves still result in a speedbump with that solution. The pod will be traveling directly toward the gap just below "seg2" if it's banking against the wall. You co…

I have a feeling that no matter how many times I'm going to answer your 'but this is impossible' with showing you how in fact it is possible you're not going to concede that in fact it probably is possible nor are you going to stop so in the interest of conserving bandwidth I'll bow out of the conversation here with this final bit.

The solution for the above is that you need to take into account the radius of curvature and you're clearly not doing that. The inside of that tube needs to be relatively smooth in the direction of travel, not perpendicular to it. You could also interleave to two segments if you wanted it to be even smoother.

So no, this is not a hard problem in spite of your continued insistence that it is because you can't solve it.

http://jacquesmattheij.com/you+can+not+do+that

Re: SpaceX: “Close, but no cigar. This time”

#232

Post by schiffern » Sat, Jan 17, 2015, 2:11 AM UTC

sillysaurus3 wrote: post #228

schiffern wrote: post #227

Earlier quoted context omitted.

> are you sure that thermal expansion imparts enough force against 175 miles of steel in order to push it all the way to the station? Nope. That's why you may need to push it! :) Assume the rollers have the Crr of a passenger train wheel (0.0020). If each section weighs 193 tonnes (23 mm thick by 2.23m diameter by 100' by 8 g/cm³), then each roller will provide ~400 N of friction. Over 175 miles (worst case) that wor…

Sweet calculations! Thanks! Check out my above comment again, since I was apparently writing out an "EDIT 2" while you were doing these awesome calculations. Basically, what do you think about the fact that it's not a long, straight tube? The rollers would need to dynamically bend the steel tube in order to carry the expansion all the way to the station, wouldn't they? Is that possible? Even if there aren't any rolle…

Heh, I love that we thought of two different but related problems simultaneously!

First yours, since it's much cooler. The minimum curve radius is 3.67 km, or a length difference of 18 mm between the left and right side over the length of one pylon.

Assuming that all sections start as straight tubes, how much sideways force would that apply on the pylon? For that we need to calculate the "flexural stiffness"[1], which for a hollow cylinder is given as:

EI = E pi/4 (D⁴ - d⁴)

Where E is Young's modulus, and D and d are the outer and inner diameter.

Therefore the bending moment[2] M is:

M = EIκ = E pi/4 (D⁴ - d⁴)κ

Where κ is the curvature, equal to 1/r.

Finally, muddling through the equations on the bending moment article (and treating the tube as a 200' beam with a load in the center) yields:

f = M / (100')²/(200') [not sure about this one]

A whopping 300 tonnes•f of lateral force on the pylons! Probably doable, but it means if your tubes are constructed straight your pylons are going to be buttressed at a ~60° angle. Also, note that this only applies to the slip distance (~150m, or three pylons), since the tube sections well within the curve will never move enough to be in a non-curved section of track, so they can be built curved from the beginning.

Also you can vary the curvature smoothly[3], which can substantially reduce the forces involved. For example, if you eased into the curve over 30 pylons and it only moved by 3, the sideways force would be reduced by 10x (since the curvature is always "about right"). Further calculations here are left as an exercise to the interested reader. :D

As for my own objection… Even on the sharpest turns successive pylons will only vary from straight line by 0.15°, so the lateral forces caused by expansion and contraction shouldn't exceed 9 tonnes force. Piddly by comparison.

[1] http://www.explorelifeonearth.org/cursos/CB-18%20simple%20st...

[2] https://en.wikipedia.org/wiki/Bending_moment#Example

[3] i.e. the curve doesn't start suddenly, but gradually; this is how actual tracks/roads are laid out. https://en.wikipedia.org/wiki/Euler_spiral

TL;DR as long as you ease into the curves (which is good for comfort anyway), it looks totally doable even without motorized rollers.

Re: SpaceX: “Close, but no cigar. This time”

#233

Post by sillysaurus3 » Sat, Jan 17, 2015, 2:13 AM UTC

jacquesm wrote: post #231

sillysaurus3 wrote: post #230

Earlier quoted context omitted.

That solution is interesting, but does it account for the fact that the tube curves all throughout its route? Some parts need to be curvy, some need to be straight. Are you sure that expansion can be coaxed into a curve shape rather than linear? How? Also, curves still result in a speedbump with that solution. The pod will be traveling directly toward the gap just below "seg2" if it's banking against the wall. You co…

I have a feeling that no matter how many times I'm going to answer your 'but this is impossible' with showing you how in fact it is possible you're not going to concede that in fact it probably is possible nor are you going to stop so in the interest of conserving bandwidth I'll bow out of the conversation here with this final bit. The solution for the above is that you need to take into account the radius of curvatu…

If you don't have time, fine. If you don't want to talk, fine. But I have some strong words which, out of respect for HN's guidelines, I won't say. Being dismissive of a newbie's concerns while killing their interest in the field isn't great, especially after they've put hard work into giving you a clear-cut example of the technical reasons why something looks hard to them.

And I still don't understand how your solution wouldn't leak air into the hyperloop unless the steel sleeves were pressing very firmly into each other. Like the inner sleeve forcefully pressing into the outer sleeve.

I have no idea why you refuse to admit this is at the very least a nuanced problem.

Re: SpaceX: “Close, but no cigar. This time”

#234

Post by Florin_Andrei » Sat, Jan 17, 2015, 2:16 AM UTC

Symmetry wrote: post #189

Florin_Andrei wrote: post #110

Earlier quoted context omitted.

Hm, if I was the guy designing the control loop on that thing (not gonna happen in real world, but bear with me), I would definitely take into account the fins position - the real position, not the one I want them to be in. Not that it would have helped anyway in this case.

I suspect that the code takes the fin's starting position into account in each planning cycle but also assumes that the fin will move as directed in the future.

No rotary encoders on those things?

Re: SpaceX: “Close, but no cigar. This time”

#235

Post by murtza » Sat, Jan 17, 2015, 2:19 AM UTC

nemothekid wrote: post #215

murtza wrote: post #184

Earlier quoted context omitted.

I am working on a project to define phrases in English like "close, but no cigar". Here's the page for this phrase: http://jargonism.com/words/437 I would be interested to hear feedback from HN on what would be useful on the word page, in addition to a definition and an example?

Why do I need to sign up and confirm a newsletter to view the site?

You are right. It is not a good user experience. I just took off the newsletter sign up.

Re: SpaceX: “Close, but no cigar. This time”

#236

Post by jacquesm » Sat, Jan 17, 2015, 2:22 AM UTC

sillysaurus3 wrote: post #233

jacquesm wrote: post #231

Earlier quoted context omitted.

I have a feeling that no matter how many times I'm going to answer your 'but this is impossible' with showing you how in fact it is possible you're not going to concede that in fact it probably is possible nor are you going to stop so in the interest of conserving bandwidth I'll bow out of the conversation here with this final bit. The solution for the above is that you need to take into account the radius of curvatu…

If you don't have time, fine. If you don't want to talk, fine. But I have some strong words which, out of respect for HN's guidelines, I won't say. Being dismissive of a newbie's concerns while killing their interest in the field isn't great, especially after they've put hard work into giving you a clear-cut example of the technical reasons why something looks hard to them. And I still don't understand how your solut…

> If you don't have time, fine. If you don't want to talk, fine.

I'm not on your payroll and you're stuck in a groove.

> But I have some strong words which, out of respect for HN's guidelines, I won't say.

Feel free, I'm not made of sugar. But you're arguing from a position that makes no sense to me, you're apparently entirely ignorant of the basics of mechanical engineering and you're going to tell the likes of Elon Musk and his band of merry engineers what they can or can not do. They just came within a hair of planting a friggin' rocket on a barge from space. Do you really believe they couldn't get a bunch of tubes to slide inside each other without overshooting their air intake budget? That's mostly a factor of 'how big are the pumps' and 'how many such sliding spots do we need' (a few hundred would do just fine) and how much room there is for additional seals (plenty).

> And I still don't understand how your solution wouldn't leak air into the hyperloop unless the steel sleeves were pressing very firmly into each other. Like the inner sleeve forcefully pressing into the outer sleeve.

No, you don't need that you can use regular seals for that, just like we do on high pressure air connections, fun fact, they work better when the difference is higher, go figure.

> I have no idea why you refuse to admit this is at the very least a nuanced problem.

Because it is a nuanced problem for you, but not for anybody that has the wherewithal to be hired on the engineering parts of a project like this. Just like my little brother would find assembly language a 'nuanced problem' and I find it relatively easy. It all depends on whether or not you're doing this (mechanical engineering) on a daily basis or whether you want to be back-seat driving nay sayer. Once you've decided on the latter any solution will be inadequate.

Re: SpaceX: “Close, but no cigar. This time”

#237

Post by m_mueller » Sat, Jan 17, 2015, 2:28 AM UTC

grecy wrote: post #2

Interesting. It had quite the lean, And close to zero vertical velocity. Then it looks like it accelerated at the last second in an attempt to move sideways back the center of the barge. I wonder if the top of the rocket was at or very near the center? Also you can clearly see the landing legs deployed, jeez they look tiny. Must just tweeted "Next rocket landing on drone ship in 2 to 3 weeks w way more hydraulic flui…

Maybe they could improve the failure behaviour with some failsafe mechanism? *If the rocket is not within a certain envelope, gain as much distance to the pod as possible? Later when they want to have landing on land, a designated area for aborting could be used.

Re: SpaceX: “Close, but no cigar. This time”

#238

Post by sillysaurus3 » Sat, Jan 17, 2015, 2:38 AM UTC

jacquesm wrote: post #236

sillysaurus3 wrote: post #233

Earlier quoted context omitted.

If you don't have time, fine. If you don't want to talk, fine. But I have some strong words which, out of respect for HN's guidelines, I won't say. Being dismissive of a newbie's concerns while killing their interest in the field isn't great, especially after they've put hard work into giving you a clear-cut example of the technical reasons why something looks hard to them. And I still don't understand how your solut…

> If you don't have time, fine. If you don't want to talk, fine. I'm not on your payroll and you're stuck in a groove. > But I have some strong words which, out of respect for HN's guidelines, I won't say. Feel free, I'm not made of sugar. But you're arguing from a position that makes no sense to me, you're apparently entirely ignorant of the basics of mechanical engineering and you're going to tell the likes of Elon…

I don't know if you're having a bad day or have gotten into a strange mindset, but I just never expected this level of "personally insulting" out of Jacques Mattheij, especially against someone who freely admitted at the very top of the thread that I'm an idiot who is arguing from a position of ignorance. So I'll go try to find someone else to geek out with an interested newbie. Like, say, someone who finds engineering enjoyable.

Best of luck to you. Keep making awesome things.

Re: SpaceX: “Close, but no cigar. This time”

#239

Post by NamTaf » Sat, Jan 17, 2015, 2:41 AM UTC

grecy wrote: post #210

vladimirralev wrote: post #202

Earlier quoted context omitted.

Not aeronautics, but control systems engineer and I've seen the explanation. The rocket engine they use is too powerful to hover and move towards the destination in a slow adjustment loop. And there is no throttle control on the engine, 70% minimum which is still too powerful. In order to land you turn the engine off and wait to pick up some vertical speed, then in the last N seconds you turn the engine on to kill th…

Thanks, that's very helpful. > The rocket engine they use is too powerful to hover and move towards the destination in a slow adjustment loop. And there is no throttle control on the engine, 70% minimum which is still too powerful I've seen this a lot, and I've also read about how many times each engine can be re-lit. Can you talk a bit about why the engines have a limited number of times they can be lit? If there we…

One limiting factor of reignition is the mechanism for lighting it. Some for example use a very combustable fuel that ignites in the presence of air to then ignite the main fuel flow. Depending on what mechanism the Merlin uses, this could limit their number of reignitions.

Re: SpaceX: “Close, but no cigar. This time”

#240

Post by jacquesm » Sat, Jan 17, 2015, 2:43 AM UTC

sillysaurus3 wrote: post #238

jacquesm wrote: post #236

Earlier quoted context omitted.

> If you don't have time, fine. If you don't want to talk, fine. I'm not on your payroll and you're stuck in a groove. > But I have some strong words which, out of respect for HN's guidelines, I won't say. Feel free, I'm not made of sugar. But you're arguing from a position that makes no sense to me, you're apparently entirely ignorant of the basics of mechanical engineering and you're going to tell the likes of Elon…

I don't know if you're having a bad day or have gotten into a strange mindset, but I just never expected this level of "personally insulting" out of Jacques Mattheij, especially against someone who freely admitted at the very top of the thread that I'm an idiot who is arguing from a position of ignorance. So I'll go try to find someone else to geek out with an interested newbie. Like, say, someone who finds engineeri…

You're not being insulted, you merely are trying to argue about things that you admit that you have no clue about and you persist in doing so regardless of the effort made to show you that you are wrong. So either stop arguing (I do know a thing or two about mechanical engineering) or start researching what I gave you above as the bits and pieces of the solution.

What else do you want? A blueprint? A cad design?