Jump to content

SpaceX Discussion Thread


Skylon

Recommended Posts

37 minutes ago, DerekL1963 said:


So long as it's less than 1/2 flights inappropriate.  If it's 1/100 inappropriate, they may not have seen it yet (or if they've seen it they may not have realized the seriousness).  Or, to put it another way, it was fine to fly with leaky o-rings - until suddenly one day it wasn't.  In the same way, it was safe to fly with substandard and misspecified parts...  until one day it wasn't. 

And while airframe manufacturers made some errors that some people lived through - they also made errors that people died from.  It's disingenuous at best to leave out that half of the story.

Well they have yet to fly people. Seriously dude, Space X is planning to send people to Mars. I would much rather be in a dragon on a F9 with a part that faults only a few seconds before S1/S2 separation at 6 g (and Im, at max, going to experience 2 g) than to be on a 36 month voyage were no significant details on how im going to be landed on Mars or return. ROFL. You guys have a hilarious lack of perspective. Think about the magnitude(s) differences in risk. If a part faults at 6-g of force and its going to experience less than half of that, thats acceptable in my book.

The challenger disaster was a completely different issue. A liquid rocket is essentially a burner that struggles to be a bomb. A SRB is essentially a bomb that struggles to be a burner. Its not a matter of if the most refined version will go wrong, its a matter of when. The problem with the O-ring on the shuttle was something the that engineers were freaking out about on the launch the very day the orbiter blew up.

CRS-7 was a flight that could have achieved its goals if allowed, S1/S2 separation was complete, and I have to say when I saw the earlier flight  telemetry.

It bothered me more that they were pushing 4+ g forces than it bothers me now that the helium storage tank broke free and ruptured. The reason for that is there are variances is fuel use, if you are pressing 4 g before flight ends and you cut your engines 1 or 2 seconds late, you are already in the acceptable g-force ranges, if you back-off the engines between 2.5 and 3.5 gs then you have a wider safety margin.

 

Link to comment
Share on other sites

23 minutes ago, IncongruousGoat said:

Regardless of how we feel about the cause of the failure, what'll be interesting is how SpaceX reacts to this new report. Either they decide the conclusions they came to in 2015 were correct, implying that they trust their own judgement more than NASA's, or they heed NASA's conclusions, implying that they trust NASA more than their own engineers. Either way, it has serious implications for how SpaceX are going to present themselves in the coming months and/or years.

If NASA has a problem, they'll not mate their payloads to the vehicle, right? I suppose that tells us what we need to know. CRS-14 is still April 2, TESS is April 16th, so they must not be terribly concerned (ie: they find the solution applied acceptable).

Edited by tater
Link to comment
Share on other sites

22 minutes ago, IncongruousGoat said:

Regardless of how we feel about the cause of the failure, what'll be interesting is how SpaceX reacts to this new report. Either they decide the conclusions they came to in 2015 were correct, implying that they trust their own judgement more than NASA's, or they heed NASA's conclusions, implying that they trust NASA more than their own engineers. Either way, it has serious implications for how SpaceX are going to present themselves in the coming months and/or years.

Good point, If the marketplace stays robustly profitable they probably will comply given their current profit margins. But I still like their model, don't overengineer to the point you are chargin 400,000 dollars to get 80t into orbit.

Link to comment
Share on other sites

8 minutes ago, PB666 said:

CRS-7 was a flight that could have achieved its goals if allowed, S1/S2 separation was complete, and I have to say when I saw the earlier flight  telemetry.


STS-51L was a flight that could have achieved it's goals if allowed.  STS-107 was a flight that could have achieved it's goals if allowed.

Or, to put it another way, that's an incredibly silly thing to say.

Link to comment
Share on other sites

SpaceX’s response seems rational to me. There’s a reason aircraft have maneuvering speed (where any input to the controls below such cannot overstress the airframe) and never-exceed speeds. Every design is a compromise. If the solution (and like @tater pointed out, likely a stop-gap solution), is simply to curtail the flight envelope, I don’t see an issue with that. Cessna knows right where yanking back on the yoke will cause catastrophic failure of the airframe. They don’t make the airframe crazy strong to survive this, they strongly advise pilots not to do it. 

It hells that SpaceX is operating their own product. 

Link to comment
Share on other sites

23 minutes ago, DerekL1963 said:


STS-51L was a flight that could have achieved it's goals if allowed.  STS-107 was a flight that could have achieved it's goals if allowed.

Or, to put it another way, that's an incredibly silly thing to say.

Revisionist : according to the report the folks at MT were anticipating failure at launch, and surprised that it looked like it would succeed.

You are drawing on emotional argument. The reality is that NASA is as responsible for the failure of STS-107 as the design of the orbiter itself.

Its not silly to argue that the American taxpayers are being ovrercharged on overengineered systems that are underperforming with respect to time. Its not silly to say that some evolution of the system needs to occur, it is expected that their will be failures it is through failures that companies learn. SpaceX appears to be learning and evolving faster then their competitors. The silly ones are ULA and ArianneSpace.

 

 

Edited by PB666
Link to comment
Share on other sites

31 minutes ago, PB666 said:

You are drawing on emotional argument.


Well, fact based engineering discussions aren't getting the point across so I thought I'd emulate your argument.
 

33 minutes ago, PB666 said:

Its not silly to say that some evolution of the system needs to occur, it is expected that their will be failures it is through failures that companies learn.


Nobody has claimed that saying such a thing is silly, nor that failures are unexpected.  So...  I'm not exactly sure what you're talking about here.
 

Link to comment
Share on other sites

1 hour ago, DerekL1963 said:


Well, fact based engineering discussions aren't getting the point across so I thought I'd emulate your argument.
 


Nobody has claimed that saying such a thing is silly, nor that failures are unexpected.  So...  I'm not exactly sure what you're talking about here.
 

My argument was not emotional, I have studied the ascent data on several SpaceX launches, I was simply pointing out to you that bad flight characteristics and a stochastic defect in the part (as SpaceX) claimed it was without any-kind of nitpicking response like NASA found. Of course SpaceX is at fault for the mistake, that's not the point, the point is that perfection is very expensive to the point we stop going to space and we let the Russians fly our boys for us. Thats not the way Americans should be playing the game we should be taking our fair share of risks too.

"Nobody has claimed. . . ." you were essentially saying that SpaceX's solution and my defense of what they did was silly, its the same as big space should win and low dollar space should lose. If the Chinese aren't in space and they want to go to space a few rockets are going to fall on villages. I bet if you allowed the inspectors from NASA go through all the components of a rocket before the Chinese launched they would have a long dirty laundry list of problems . . . . .if you want to go to space you got to be willing to take a few knocks.

Link to comment
Share on other sites

44 minutes ago, PB666 said:

the point is that perfection is very expensive


Nobody was claiming that things had to be "perfect".   Choosing the proper material for an application is neither nitpicking nor perfection - it's basic bog standard engineering practice.
 

46 minutes ago, PB666 said:

"Nobody has claimed. . . ." you were essentially saying that SpaceX's solution and my defense of what they did was silly, its the same as big space should win and low dollar space should lose.

No, I was not essentially saying any such thing.  (Though your wildly disconnected from reality defense of SpaceX's failures was indeed very silly.)

And speaking of emotional arguments....  wow.  Just, wow.  I'm done here.  You're not interested in anything that isn't blind worship of SpaceX.

Link to comment
Share on other sites

8 hours ago, tater said:

I'd add that we don't know all the Block 5 changes (and that's the F9 that crew will ride). Perhaps (likely, IMO) they include all the changes already recommended in the report.

According to NasaSpaceflight.com, the changes were made in time for the Jason-3 mission. Except possibly for the telemetry, although they speculate that that has probably been addressed too, given that SpaceX has now received Category 2 certification from NASA.

Quote

The IRT report finished by noting that all of the key findings in the report were addressed by SpaceX in time for the successful Jason-3 mission for NASA.

However, that launch utilized a Falcon 9 version 1.1 vehicle. The “Full Thrust” variant is version 1.2, so it remains unclear if the telemetry issue has been resolved.

That being said, in January of this year the Falcon 9 Full Thrust received certification to launch Category 2 NASA missions. Category 2 certification allows Falcon 9 to fly medium value NASA payloads. Such certification would not have been warranted if NASA did not have significant confidence in the Falcon 9 launch vehicle.

 

Edit:  The article is worth reading and paints a rather less scathing picture of SpaceX than the response to that tweet. Yes they screwed up, but their post-flight investigation was quick, seemed to be robust (in that NASA independently came to the same conclusions) and, as mentioned above, the key findings in the NASA report were addressed promptly and with little to no fanfare that I remember. 

Edited by KSK
Link to comment
Share on other sites

Well, the article cites NASA saying that SpaceX used a part without testing and outside of manufacturers specifications.

“the implementation was done without adequate screening or testing of the industrial grade part, without regard to the manufacturer’s recommendations for a 4:1 factor of safety when using their industrial grade part in an application, and without proper modeling or adequate load testing of the part under predicted flight conditions. This design error is directly related to the Falcon 9 CRS-7 launch failure as a ‘credible’ cause.”

"In simpler terms, the steel strut that SpaceX chose was not certified to be used in such conditions. Furthermore, SpaceX did not meet the 4:1 redundancy requirement that the manufacturer had instructed."

 

So, it was not a manufacturing defect but bad engineering. The strut actually held what it was designed for.

The article also says that SpaceX corrected that immediately after being informed by NASA. I can imagine that a certification would have been revocet immediately had they not done so. SpaceX informed the public differently, by stating or implying that the they changed the manufacturer because of bad material, which is not the case.

Furthermore:

"That being said, in January of this year the Falcon 9 Full Thrust received certification to launch Category 2 NASA missions. Category 2 certification allows Falcon 9 to fly medium value NASA payloads. Such certification would not have been warranted if NASA did not have significant confidence in the Falcon 9 launch vehicle."

iow: SpaceX had to change it because their rocket would not have been certified.

 

Anyone remember SpaceX suspicion that somebody could have shot at their rocket when it popped on the pad ? Haha ... :-)

 

In the end it is not that bad to always have an independent investigation on complex failures. It might help to find out about the real causes and avoid them next time.

 

Link to comment
Share on other sites

Agreed, @Green Baron but it's also water under the bridge now.  Spaceflight is very good at exposing organisational weaknesses and is littered with incidents caused by intelligent people (this is rocket science after all :) )having a bad day. That bad day might be caused (for example) by inadequate regard for a manufacturer's recommendations, placing a little too much faith in your Russian sourced engines, messing up a 64 bit to 16 bit data conversion, installing sensors upside down, not fully appreciating the fire hazards involved with 100% oxygen atmospheres or mixing up imperial and metric units. Unfortunately, rockets don't much care about the human factor - they just go boom anyway.

Speaking as a SpaceX fan, it's obviously disappointing when they do make mistakes (which CRS-7 clearly was) but, on the other hand, there's no reason why they should be immune from mistakes either. They have good people but so do Orbital, ESA, Roscosmos and NASA. No doubt Blue Origin, Rocketlab and the rest will have their share of bad days in future. Virgin Galactic have already had their share.

 Not an excuse - just an observation.

Edited by KSK
Link to comment
Share on other sites

So far as I understand, the strut theoretically had the capacity to hold that force, but had not been rigorously tested under cryogenic conditions, and SpaceX's usage of it didn't meet the manufacturer's suggestion of a 4:1 safety factor. I'm no engineer: I suspect in non-critical applications, a 4:1 safety factor is very... safe, but when failure endangers a rocket costing tens of millions of dollars, well...

Well, in that case, using an industrial-grade strut end without much safety factor destroys your tens-of-millions-of-dollars rocket. You might get away with it initially, but if there is even a minor manufacturing defect, well, now your rocket goes boom.

What's a little bit more troubling to me is the comments about the telemetry. It's not cheap or easy to redesign a telemetry system; while mechanical parts, you can sweep through and replace them with better-rated parts when you realize you've been trying to cut too many corners, redesigning a system intended to talk with every part on your rocket is a very nontrivial task. Assuming they haven't already done it, it's not going to be cheap or easy to ensure their telemetry system is up to NASA's standards.

Link to comment
Share on other sites

5 hours ago, Green Baron said:

The article also says that SpaceX corrected that immediately after being informed by NASA. I can imagine that a certification would have been revocet immediately had they not done so. SpaceX informed the public differently, by stating or implying that the they changed the manufacturer because of bad material, which is not the case.

 

Which is why I made the point in my first response that Space X was pushing high g-forces in their early flights. And it my second response I made the point that once the craft is traveling with a significant horizontal component that it was unnecessary to push 3.5gs let alone 4 something g-forces. When you talk about 4 fold tolerances, that includes the bump g-forces in a rocket. You can see the bump data here. The orange lines are refactored bumps. Note the magnitude of the bumps with respect to TWR, they, at max run about a quarter of TWR.

LQpxPOH.png

As the first stage gets lighter those bumps become more severe as the inertia of the first stage is absorbing less and sending more to the second stage. This data is not as bad as what it in the telemetry, the telemetry shows more forces.

Here is the delta acceleration (m/sec3) on a per second time stamp for falcon heavy approaching the 187 second S1-core shut down. We have to remember that the core is still carrying alot of fuel, so this is less than the bump data for what an empty core would be prior to shut down. Since time stamp is valid to within about 1/25th of a second and the telemetry data is not pure there are other measures in the data. It does not capture the peak or trough of the bumps either, the final two anomalies are expected, as the engines were shutting down. The data is actually more bumpy but timestamps of half or quarter second cannot be validated.

  1.3
-7.7
  5.9
  0.5
  1.6
-3.7
 0.3
 1.3
 2.4
 1.8
-4.7
  0.5
  4.5
-1.1
-3.4
  4.2
-17.1
-5.3

Easily you have ±1 g/s delta-accelerations when engines are runnning at full thrust which can be added to whatever acceleration the rocket is experiencing, and easily pushing moments of 4 g loading. Since the F9 has lately generally been RTLS or DSL their tanks have never been close to empty, keeping 10% of the fuel in the tank. With a load of 110 t + 41t BB fule and emptyS1 weight of 22t the fuel adds 31% to the weight of the ship that is not present in a fully expendible first stage. Consequently the bumps one sees only get worse as the S1 is burned to empty. Accelerations increase and bumpiness increases.

SO the problem was precisely defined by both groups as this, you have a tank within a tank. One tank (pressurized Helium) is much lighter than the other (Oxygen), it wants to float. The higher the g-force, the more it wants to float. One that launch pad the force is k*1g. In early flight the force ranges from k*1g to k*1.5g. After MaxQ the force begins to increase, if your engineering tolerance is 2g with 2 fold safety factor, if your g-force exceeds 3g then the bumps place you in the 4 g range. But telemetry data from about 15 flights back suggested they were bumping up to between 5 and 6 g-forces. You are putting every part in the craft, from stage fairings to S1 upper structure at risk of failure, unnecessarily because.

1.  your drone ship can wait any distance off shore, there is no reason to be close as possible to the launch site.
2. If you held g-forces to below 3.0 you are really not loosing that much dV, less than 0.25%.
3. its a very brief period of time.

On that day that part failed, with the right bump on another day, another part may fail.  Then the investigation would go back and find some other part was (enter bureaucratic reasoning). It makes no sense to improve every part on the craft when just reducing thrust simply fixes the problem. The reason the part failed was very simple, while it may or may not have had a defect, at the end of S1 flight the rocket is pushing its highest g-forces, that 3 seconds before power down was when the rocket reached maximum g-loading, thats the moment it failed, that is why the part failed. For every action there is an equal and opposite reaction, for every force applied there is an equal and opposite force. Parts do not break without the application of force. Lowering the forces and the accompanying bump also lowers the risk of other parts failing. Its simply a matter of powering the engines down by 20 to 30%.

If I take your car and place 6 roof storage boxes on the car its not the fault of the manf of the roof or the boxes _if_ the roof fails. If I take your engine on your car and using nitromethane run the pressure of cylinder 2 past the maximum pressure, its not the fault of the engine if it fails. If I have a skiier on the back of a boat, and I place a motor overrated for the boat and tear the rope attachment off its not the skiers or the rope attachments fault.

My observation about SpaceX is this, they make changes, often expecting failure (which they often state in advance as a likelihood), if it doesn't fail they move forward, if it does fail then they make small modifications. This keeps the cost down, by simply altering their ascent dynamics they avoid future problem, a small change fixes the problem, then they should go with it. If the customer wants to pay for something more, then let them pay. But don't raise the price for all customers simply because one customer wants to burn at maximum g-forces.

Note: if anyone wants to see the unvalidated data, I can do this for any flight that has telemetry data, for time window of a few seconds.


 

 

Edited by PB666
Link to comment
Share on other sites

2 hours ago, Starman4308 said:

So far as I understand, the strut theoretically had the capacity to hold that force, but had not been rigorously tested under cryogenic conditions, and SpaceX's usage of it didn't meet the manufacturer's suggestion of a 4:1 safety factor. I'm no engineer: I suspect in non-critical applications, a 4:1 safety factor is very... safe, but when failure endangers a rocket costing tens of millions of dollars, well...

Well, in that case, using an industrial-grade strut end without much safety factor destroys your tens-of-millions-of-dollars rocket. You might get away with it initially, but if there is even a minor manufacturing defect, well, now your rocket goes boom.

What's a little bit more troubling to me is the comments about the telemetry. It's not cheap or easy to redesign a telemetry system; while mechanical parts, you can sweep through and replace them with better-rated parts when you realize you've been trying to cut too many corners, redesigning a system intended to talk with every part on your rocket is a very nontrivial task. Assuming they haven't already done it, it's not going to be cheap or easy to ensure their telemetry system is up to NASA's standards.

See below, you don't need perfect information, but it would be nice to have shake data on a number of key points, my data is lousy, but even lousy data shows siginificant shake, that data shows that as accelerations go up, so does the bumpiness.

Note:   Temperature of liquid O2 is −182.962 'C, the steel used was a precipitation hardened steel of "SAE type 630 stainless steel (more commonly known as 17-4; also known as UNS 17400 [1] ) is a grade of martensitic precipitation hardened stainless steel. It contains approximately 15-17.5% (17) chromium and 3-5% (4) nickel, as well as 3-5% copper." wikipedia (C <0.07%, Mn <1.2%, P <0.04%, S <0.03%, Si <1%, Cr  15-17.5%, Ni  3-5%, Cu  3-5%, Nb+Ta 0.15-0.45%
 

Quote


Low carbon 3.5% nickel steel is widely used in large land-based storage tanks to contain liquefied gases at temperatures down to -150'F [-100'C].
This steel falls under ASTM A203, Grades D and E, and is subject to impact tests in accordance with. the requirements of A-300.https://www.bnl.gov/magnets/Staff/gupta/Summer1968/0311.pdf

The failure between 100 to 200'C is not an all case failure but a specific type of usage failure known as "notch embrittlement" here is what it means. https://en.wikipedia.org/wiki/Charpy_impact_test#Definition

Therefore the conclusion of SpaceX is probably correct in that context, damage to the exterior of the metal could have occurred during manf, installation or flight. A perfect situation would be that something freed itself from the tank and struck the piece. However I should note that striking could also be two connected pieces vibrating against each other (see above post with graph). The vibrations late in flight would have also be adequate to notch the loop, sheering it from it mooring and freeing the tank.

Edited by PB666
Link to comment
Share on other sites

11 minutes ago, tater said:

As an aside, related to commercial crew safety requirements, I wonder if the Apollo CSM or LM would come remotely close to passing muster.

We were a different people back then.

Link to comment
Share on other sites

28 minutes ago, tater said:

True, but the Soyuz spacecraft is of the same era, and we pay to fly astronauts on it to this day...

It is ironic isn't it. Hopefully SpaceX does not change their risk reward metric and keep pushing the envelope.

Link to comment
Share on other sites

3 hours ago, Starman4308 said:

So far as I understand, the strut theoretically had the capacity to hold that force, but had not been rigorously tested under cryogenic conditions, and SpaceX's usage of it didn't meet the manufacturer's suggestion of a 4:1 safety factor. I'm no engineer: I suspect in non-critical applications, a 4:1 safety factor is very... safe, but when failure endangers a rocket costing tens of millions of dollars, well...

Well, in that case, using an industrial-grade strut end without much safety factor destroys your tens-of-millions-of-dollars rocket. You might get away with it initially, but if there is even a minor manufacturing defect, well, now your rocket goes boom.

What's a little bit more troubling to me is the comments about the telemetry. It's not cheap or easy to redesign a telemetry system; while mechanical parts, you can sweep through and replace them with better-rated parts when you realize you've been trying to cut too many corners, redesigning a system intended to talk with every part on your rocket is a very nontrivial task. Assuming they haven't already done it, it's not going to be cheap or easy to ensure their telemetry system is up to NASA's standards.

I think there was a comment in response to the original tweet that suggested the telemetry fix should be straightforward as an appropriate protocol was available off-the-shelf. Although I have absolutely no idea how reliable that source is.

As for the strut - that makes sense to me. There may well have been a defect in the strut (as per SpaceX's claim) but that wouldn't have been a problem in itself, if SpaceX had gone with the recommended (possibly slightly overkill but yeah - tens of millions of dollars at stake) 4:1 safety factor. So both sides are correct -  from a certain point of view (</obiwan>)

As usual, it's never just one thing that gets you. Or, as @DerekL1963 pointed out - everything can look fine, right up to the point when it isn't.

Link to comment
Share on other sites

Okay, I could have sworn that, around the time the initial SpaceX report came out, I heard that the strut that failed was rated for 5,000 lbf, and failed at 1,000 lbf, which is a 5:1 safety factor. Does anyone have any hard figures for the rated load and actual load for that strut?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...