SpaceX Explosions - Engineering Done Right
Вставка
- Опубліковано 19 гру 2023
- A discussion of SpaceX explosions and why they aren't a problem, but instead are a strength.
Ignition! book is here: library.sciencemadness.org/li...
@Eager_Space on Twitter
Triabolical_ on Reddit
/ eagernetwork
/ eager-space-1038430522... - Наука та технологія
Explosions are good. Learn it. Love it. Live it. AKA How I learned to stop worrying and love SpaceX.
I really considered "How I learned to stop worrying and love SpaceX explosions", but I've done the Dr. Strangelove reference in the past so I decided to do a different one.
@@EagerSpace😂 such a good reference
Dude... I loved that. "This marks the first time spacex achieved two explosions in the same flight"
Thanks.
I am so thankful for your insightful videos and the strategic educational value they provide me as product planner. Alone in the office one is often struck with the sense that your challenges are unique because you are working on new technologies when the reality is quite the opposite. There is much to learn from the past and I find myself encouraged with every episode.
Thanks for your kind words...
Before I retired I spent a career working on software where "explosions" are really par for the course, and I learned to really like the agile "fail fast / try again" approach over the waterfall "build something big and ship it with a ton of known bugs" approach.
I've also paid a fair attention to lean and the theory of constraints.
The only risks of this method is SpaceX having free-rent in certain public head & they started barking that the first launch must be successful no matter what
The only time first flight must be a success was the Appolo program. Maybe that had built a wrong expectation for a lot of people.
These are the same fools who then wonder why development costs have exploded. If you don't allow engineers to do destructive testing then they have to resort to simulations. For some reason people think that computer time is free when the reality is you need very powerful computers to run your simulations fast enough for the data to be meaningful. Even then the old Garbage In Garbage Out rule still applies, the simulation is only as good as the programmers who created the simulation.
Thank you for creating this exposition on the method SpaceX is following, one that hasn't been used in decades and is thus not understood by most people, even some who follow spaceflight forums.
Looking back at the timelines surprised even me, who's been following Starship since there was a single tent in Boca Chica. It took only 2 years to go from Starhopper to SN15's fully successful flight and landing!
Glad it was helpful!
Here I'm gonna disagree with Starship being less uncharted territory then F9 reuse.
It's not just landing an orbital class booster back, it's having a completely different flight profile from F9, with continuous engine burn, hotstaging, different return maneuver AND, mostly importantly, it's not just landing a booster, it is turning the biggest airborne vehicle ever to the launch site and CATCHING IT with a massive static tower. And then, of course, we have Starship reuse that is making a completely new maneuver to "land" on that same chopsticks after reentry. That is NUTS!
I'd say Stage 0, 1 and 2 reuse are more uncharted territory than F9 stage 1 reuse.
Yeah, return to base being the only option for the booster AND the cargo vehicle multiplies risk of failure. Fortunately, if anyone has the required expertise to pull it off it's SpaceX.
There's obviously room for discussion on that point.
Part of my criteria for "is this hard" is whether SpaceX has bothered to test it separately. They very easily could have tried a booster hop to do the F9 equivalent test, though Raptor supply might have been an issue. But they decided not to bother - they're going to practice with controlled water landing and then probably go for the tower catch.
Falcon 9 took a while to get those to work, but SpaceX has all the starship engine testing to look at, and booster's landing requirements are a lot easier than starships.
I previously thought that second stage reentry was going to be a really big challenge, but these days I'm more optimistic. There's a ton of data from shuttle and Apollo and SpaceX has their own data from the Dragons. I expect it to be more development than R&D.
That's all without mentioning that it's the first every operational rocket to use both methane and full-flow staged combustion! Raptor alone is a marvel of new engineering
@@SuperSMT full flown engines actually never flew before, so that's a big new thing in starship
Agreed. Starship involves a brand new stage 1 and a brand new stage 2, unlike other examples where only one stage is being developed. That alone will double the unknowns and earn the "two explosions in one day" achievement. :)
The airplane example is such a great explanation of why SpaceX's approach makes sense. It's already proven as a better approach. Great job pointing this out.
You have a new susbcriber - thanks to @KenKirtland17 for retweeting your original tweet for this video. If Ken is enthusiastic, then everyone else should be too.
Thanks. I'm happy for Ken's support.
It's cool to think that falcon 9's re-entry method came about because the other method wasn't working. A company that thinks themselves above testing might be spending to much time and effort optimizing for something that is far from effective.
They are very much about testing something to see how it works and adding in complexity only as necessary.
This is really clear with starship.
I'm amazed at how many people don't get the MVP/prototype-hardware rich approach SpaceX are going for and laugh or deride Starship for all the explosions, like they just don't either understand the point of prototype development testing and lack the mindset to understand it or more likely they don't follow rocketry that closely so for them NASA and their approach of "get it right the first time" is what they expect it to look like especially when you have a vehicle as massive, complex and ambitious as Starship and Super Heavy I guess.
Great video, great overview of Starship R&D and the philosophy behind it.
I had a section on engine development that didn't make the cut...
It's been accepted for years that if you are building rocket engines you are going to have to build a lot of prototypes and you are going to have explosions, but for the rockets that you put the engines in you can get away with analysis. In some cases you can, some cases you can't.
I'm eagerly waiting to see where rocket lab ends up with their test approach for neutron.
@@EagerSpace "I'm eagerly waiting to see where rocket lab ends up with their test approach for neutron."
How compatible are explosions with the stock market? I think any publicly traded company won't fully go with such approach.
I'm more curious what Stoke will do
When I try to explain to my friends how IFT-1 and 2 were successes I tell them they were "mid-air grouped component testing", not test flights in the sense of the term used by other rocket builders.
What a great, informative video. Even as someone who already knew about all the history you shared, your delivery still demonstrated the accuracy and legitimacy required to keep my attention. This video would also be informative and entertaining for someone new to the material. Well done and thank you!
I think SN15 landed successfully
excellent video, as usual
Had an engine failure during relight, which is why it landed on the edge of the pad, but did land successfully
Great video as always man, thanks a lot. =)
All the true SpaceX and Musk fans applauded the results of IFT-1 and 2, which were about what everyone expected. Only his haters considered them failaures.
Even as a devout SpaceX fan since the beginning of Falcon 9, I think IFT-1 was at least a partial failure. Elon's stated goal was to get off the pad without destroying it.. they did get off the pad, but kinda destroyed it in the process!
IFT-2 on the other hand was a resounding success
@SuperSMT Fair enough. The concrete tornado was definitely a setback. Still, all part of the process. In order to reduce costs and expedite production, you have to experiment with ways to save money. Nobody knew for sure whether the concrete pad could survive the thrust. By not installing a water deluge system, they saved tens of millions and were able to launch sooner. Unfortunately, it didn't ultimately spare them those costs and resulted in significant damage to other structures. As Elon has often said, the best part is no part at all. I wonder how much simpler starship, superheavy, and Starbase are because of this design philosophy.
@@SuperSMTi see IFT-1 as taking "clearing the pad" too literal.
The pad survived the lower-thrust static fire test, so there was a decent chance that it would survive IFT-1, but I think SpaceX made a significant miscalculation on the regulatory consequences that came from creating a concrete shower in a sensitive habitat.
@EagerSpace Why is Boca Chica a "sensitive habitat"? How is it any more special than any other undeveloped land? There's no objective reason to believe it's more or less "sensitive" than any other ecosystem. Sorry, personal grudge I have against environmentalism. It isn't based on any cogent philosophy as to why natural ecosystems are good or whether you can consider an ecosystem healthy or not. To me, it's obviously not truly about conserving ecology.
"Neither new nor original"
Applies in all fields. As one of my history profs used to say, "Every time I think I've come up with a new insight, some damn German wrote a paper about it in 1882."
He's exaggerating...but not by much.
Yes.
In this case I was happy to see it because it gave me support in a historical context.
Excellent
Excellent!
Many thanks!
Explosions are good: Learn it, Know it, Live it.
[knock-knock on the door] Hello, this is the ATF SWAT team. Will you kindly let us in?
Great video!
Thanks!
It has been a wild ride.
Would it be fair to say that those who mock SpaceX's R&D method are in the bottom quartile of knowledge regarding R&D methods -including much of NASA?
Somewhere I have a video I did on Dunning Kruger because I found there were so many people who were in the first quadrant for Dunning Kruger.
I think the decision for Falcon 9 reuse was really easy because the marginal cost of development was so low.
I'm impressed that SpaceX decided to continue that for Starship, where they would start with poor prototypes and then just keep iterating until they got them to be good enough. It's a bold approach, but part of the feasibility is because Musk really wants to build a factory that can build 100s or even 1000s of these vehicles. They're mostly leveraging the flight side off of test vehicles from the factory side. I'm not sure the same approach works for companies with aspirations that are less crazy.
I originally had a section of NASA and shuttle design that I cut - it might come back in another video - but comparing them to SpaceX is was really obvious that "design and produce" was one of the reasons why shuttle was such a poor vehicle - NASA had this hugely speculative architecture and a very advanced orbiter and they decided "we know how to build stuff and we can come up with the right design and not have to modify it". Epic amounts of hubris in that.
I do think that some of the skeptics have forgotten what they thought when spacex first started with Falcon 9 reuse.
The engineers at NASA and the higher-ups, even the good ones, labor under the burden of public scrutiny of spending taxpayer dollars. An explosion makes the news, just like SpaceX, but the criticism isn't just from social media Elon-haters. Congress asks questions. Even Congress members who know better don't refrain from criticism if it serves a political purpose. Publicly owned company that answer to stock holders face a similar problem. SpaceX is a complete anomaly - because Elon Musk is a complete anomaly, someone willing to spend a very large fortune on taking rocket technology several leaps forward. Nobody else has a boss who has such a high tolerance for failure, or what's usually considered failure.
Maybe they just don't believe in stupid fairytales, you ever think of that?
@@donjones4719that was a long reply that shows you don't have a clue how the world works.
@@EagerSpaceTHUNDERFOOT right here on yt does a good job of destroying the musk rats lies, by doing nothing more than using his own lies he spews against him. Might be a good watch for you dunning Kruger types!
Amazing video! Incredibly refreshing to hear some rational discussion about Starship lmao.
Thanks.
That was a very informative, and very succinctly put SpaceX's development philosophy, that Prior Art-Explosiveness chart would be extremely useful when trying to explain to friends and strangers just how cutting edge SpaceX is. With SpaceX, explosions are never failures, it only becomes a failure when nothing was learned from it to improve.
Decades of the Define and Produce in Old-Space has really mucked up the general public's expectations on Aerospace development.
As of posting this comment, the next Starship flight pair is already undergoing pre-launch testing. Just about a month since IFT-2, a very fast turnaround time, another demonstration of just how differently SpaceX works.
I've been looking for some projects that push technological innovation a lot to see what they did and there really aren't very many. Shuttle is an example of define and produce, and IMO it shows all the problems of that approach - their approach didn't allow for any time to take a step back and figure out how well the plans were working before they built the whole fleet, and that caused some obvious problems.
With Define and Produce booster landings are impossible
I've been watching so many Thunderfoot videos that I never expected any space or rocket enthusiasts to praise SpaceX. But it is a surprise worth welcoming in my opinion.
Why would a space or rocket enthusiast watch Thunderfoot videos?
Thunderfoot is quite a bubble. From what I see most rocket enthusiasts are very enthusiastic about SpaceX and Starship's potential, even in spite of Elon being so cringe online.
Thunderfoot has no idea what he's talking about half of the time when the topic is spaceflight... i wouldn't even consider him a spaceflight enthusiast.
@@TheNheg66i dare to say he has no idea what he is talking about on pretty much any topic... watched his video about hyperloop a few years back, and had to stop when he argued you have to pump the air back into the vac tube in case of an emergency which will take hours..i just thought "yeah, vacuum sucks" and stopped the vid, because there was nothing to learn from it..
Thunderfoot has no idea what he's talking about when it comes to spaceflight.
And when Starship flew regularly every week like Falcon 9 is everyone will be like:
"Well it's not impossible. If it was (other space company) they would pull it off in one try unlike SpaceX reckless decision making it explode a lot of time during development"
That's not how it works at all blud. The first animal in space disintegrated not because we're reckless but we have no idea how it works. A theory is only a theory until it was proven right in reality.
Knocked out of the park, as always.
I think we’re supposed to call the missions Starship Flight N now and drop the ‘IFT’ or ‘OFT’ lingo.
Just found your channel. You look to have a lot of great videos and I'll enjoy binging them over the next days 😉.
SpaceX is just stomping the competition. The only chance any old space company has is government welfare for multiple launch redundancy. They are all aiming their new rocket development for the same old expendable design. I can't fathom how they expect to succeed without trying for reusabulity. New space has a better chance, but they as a company are much more fragile economically. Still I have much more respect for companies like Astra and firefly
Thanks.
SpaceX was both more driven and nimble and they took advantage of the unique conditions when shuttle retired - see my recent video on how spacex had failed.
Old space did very well with their existing model - it reliably made them billions of dollars in profits. From a business perspective, ULA had no reason to try to be cheaper - yes, it would increase their profits but it would cost a lot of money and they don't fly enough to earn it back in a reasonable amount of time. Same with Ariane - a *hugely* successful business model - I hadn't realized until I looked at the history, but they came in with a new rocket and an innovative idea - launch two payloads at once - and absolutely ate Lockheed Martin and McDonnell Douglass's lunch - I was amazed how dominant they were when I did the research for the communications satellite launch wars. Falcon 9 has launched 52 comsats, proton has launched 84, Ariane has launched 387. Their big problem is a) they can't easily flex their flight rate and b) they forgot that they always overlapped their old and new rockets and have been stuck without a rocket waiting for Ariane 6.
My second place company is Rocket Lab as they are the only one flying regularly which means they are actually competing against SpaceX. And I think Neutron plays to their strengths and SpaceX's weaknesses.
I'm not excited about the other launch companies. They may be fine from a tech perspective but I see no way they can compete with Falcon 9 in terms of price.
There is the "anybody who isn't spacex" market that Kuiper has created. Though I'm not sure Kuiper is actually going to proceed - Amazon appears to have stalled and I'm wondering if they are re-evaluating their plans.
nice
The Kerbal Way
Space X should leave it to the professionals like Boeing for rocket development!😊
I've a feeling I heard a while after the SN11 flight that it suffered a hard start on the engine(s) during relight for the flip manoeuvre
Excellent video that needs to be viewed by everyone interested in SpaceX, Falcon 9 and Starship. A gigantic load of useful and important information in just 21 minutes.
Thanks, I'm glad you found it interesting.
When talking about SpaceX the video quickly gets outdated! I didn't notice the published date at first, bummed it didn't cover flights 3 and 4. Not much we can do about that whole temporal thing.
That is certainly true, especially when you're trying to talk about SpaceX.
I think the point still stands - SpaceX has made a ton of progress on flights 3 and 4.
lol at the end quote
The whole book is full of that kind of stuff...
I know its way too early to ask but once they start making changes for crew rated flight, how many fundamental changes will be needed to make the transformation? Will Starship as we know it today resemble the crew rated version? Are life support systems relatively simple to add on?
I personally think crew starship looks a lot like the current starship. Many people think it needs an escape system but I'm not convinced; if Starship can be as reliable as Musk asserts it can be, it's possible an escape system might actually make it more dangerous.
I suspect spacex is the reason why at least a few people aren't learning how to speak chinese right now.
How am I just now finding your videos???
I have Ignition, but haven't read it yet..
You must. Highly informative and entertaining.
So, the real issue with Starship is the overwhelming number of elements that have “no prior art.”
They’ve blown up 3 IFT’s and still haven’t reached orbit with a complete system (Starship itself is still an empty hull). They also haven’t successfully landed and recovered Super Heavy.
They still need to demonstrate relight of a raptor engine in space (without green flames cannibalizing itself), the ability to control Starship re-entry, actual ship to ship fuel transfer, and then they need to do all of that with Starship built for habitability, not just an empty shell. Then they need to demonstrate that they can land on the moon…without tilling over. And then launch from the moon. And successfully return to Earth.
So before a “string of successes” like Falcon 9, we should expect how many more explosions of Super Heavy, how more explosions of Starship, 2 or 3 during the “research” phase of every thing they haven’t done yet?? And then what? Human rate it, have a string of successes like Falcon 9 and then, “oops” we kill a crew after success 3? And then another crew only 28 launches later?
At the rate SpaceX is going, there are just too many elements with “no prior art” and so my prediction (as a 40 year old) is they will never get enough funding to see Starship through to reliability, and even if they did, it will never happen within my lifetime.
Starship is definitely the most ambitious rocket ever built.
The problem for us is that we don't have much data about the actual state of the program - see my "you suck at space" video - so it's not possible for us to make a rational analysis. All we're left with is opinion.
I expect future launches to be more successful but getting back through reentry is no picnic and nobody has done it with this kind of vehicle and that may take a few tries.
But I would also point out that Falcon 9 landing has been developed to the point where it is essentially perfect, and crew dragon has been pretty much perfect as well. And what you are saying about Starship is exactly what people were saying about Falcon 9 landings.
My only reason for calling out Falcon 9 is to point out, your own video rationalizes all of F9’s failures to be attributed to “no prior art” when it came to landing and recovering the booster stage…that’s it, one thing, land the booster.
So I think the video falls short of explaining the magnitude of how much “no prior art” there is for all Starship has yet to accomplish.
Maybe if your chart at 3:23 added some columns after “Lots” called “Bunches,” “Tons,” and “Butt Tons” then put Starship’s dot somewhere to the right of that, it would more accurately reflect the magnitude of what SpaceX is trying to do?
Then re-explore the “Falcon 9 blew up a lot because there was no prior art for landing” and explain what that means for how many more explosions and failures we should realistically expect for Starship.
@@The93Vector to be fair, the last 3 things on your list have been done before (get to, land on + return from the moon) in the Apollo program, so it's not completely uncharted territory. Also, in terms of the other issues, I think SpaceX's hope is that they can manufacture Starships quickly enough and cheaply enough that they can deal with a few explosions in tests. I do think getting human rated will be very hard for starship tho, at least for ascent and descent- maybe we'll see some sort of Dragon docking with Starship in LEO in order to put humans on it?
Keep in mind over a century ago people said that the concept of the plane was poorly conceived, and many people where critical of it throughout its conception and development.
Only a few decades later it became an essential part of society, and even more so today carrying immense amounts of people and cargo across an expansive range. This in comparison to the simple pleasure craft it originally was.
Planes today may seem ordinary to us, but back then they were anything but.
I’ve got a couple caveats here:
- I don’t see where the eight explosions during F9 landing development came from - they certainly had more than that, even by your accounting in this video?
SpaceX didn’t come up with the term “supersonic retropoulsion” - I think your description of the events around that would suggest they did.
Overall, I must agree with the general point, though it might have been wise to mention earlier VTOL rocket work as some prior art for F9 landing even though they weren’t specifically boosters.
8 explosions before the first landing, as shown in the video. If you want to add the 2 additional ones before the first drone ship landing, I'd be fine with that. There were more later; I think I cover all of them when I talk about the program in detail. You can get into semantics about what the true meaning of "explosion" is. Running into the drone ship is an obvious one.
I originally had a section where I talked about DC-X, but I ultimately decided that while it was prior art for grasshopper and F9R, it wasn't prior art for getting through the atmosphere or maneuvering to get to the landing spot, and that hovering and landing wasn't that hard to do. The issues that DC-X and F9R had were about auxiliary systems, not about the hovering and landing part, and neither of them had to do it on low propellant with a non-hovering vehicle.
Or, to put it another way, if you can do something new 8 times in a row it's probably not that difficult to do.
@@EagerSpace ah, that does clear it up. Thank you!
I get it , explosions are good. On the scale 1 to 10 how worried are you about the chopstick arms? Even if we assume its a good idea from long term perspective , how are they going to avoild the huge bill from those few failures ? Build some new poligon to just test them separately?
Assuming SpaceX does enough water landings to validate that they have good control (one might be enough) and they don't have engine issues or mechanical failures in the catch mechanism, I think the catch will work first time. It probably has slightly higher accuracy requirements than landing on a pad but I don't think they are a lot higher and they can hover super heavy (though they will want to limit that). The important part of the catch is that it can happen *slowly*
I should note, however, that I don't think I'm a particularly good guesser about either what SpaceX will choose to do from a test perspective or whether what the success fill be.
Even IF catching is taken off the table, they're still an important part of the infrastructure.
Sorry to nitpick, but SN8 was in December 2020, not early 2021.
Thanks for the correction
I follow Starship development eagerly and hope for the best. But the devastation to the launchpad during IFT-1 was expensive in terms of credibility due to the forseeable nature of the damage. AFAIK it was Musk deciding against better advice that they should launch with a flat piece of concrete under the stack.
All consequences it took was a few months of repairs
@@alvianchoiriapriliansyah9882 You make it sound triviel. But imo it wasn’t. A bit more damage and the stack could have toppled while still on the pad. Waving it of as nothing is not doing any good.
@@tusse67 Man I'm just glad they didn't have approval to launch Booster 4 - Ship 20 in time, I think we would've seriously gotten a RUD on the pad, and if it was anywhere else it would be fine overall but considering the site its on, the environmental and political fallout BS involved might've come down even worse on SpaceX than for IFT-1
I don't think the damage to the pad was a significant issue; they were able to repair it fairly quickly.
I do think they underestimated the amount of attention the failure would bring. It seems likely that delayed IFT-2 and it's possible that there are more issues to come.
@@tusse67 Musk deciding against advice? Can you cite and credible source for that? So far as I know that's just pure speculation.
Starship exploded because they vented the spare oxygen.
All nice and fine with SpaceX iterative design, but they aren't on the end of the road, Musk itself said that only version 2 can be useful for delivering real payloads, and those would require new materials and engines, how much would takes until a real operational vehicle would fly? Nobody knows!
I'd have to see the quote and the context to have an opinion...
The current version likely has a 100 ton payload to LEO. The "coming soon" updated version with a stretched starship and other unknown changes is probably good for 150 tons. I think that new version will probably fly next year.
SpaceX would really like to be launching Starlink V2.0 satellites with Starship as soon as possible, and I think 2024 is a reasonable date for that first flight.
@@EagerSpace not a single one of the existing version prototypes would launch any significant payload , I said that long time ago but I think that thing is obvious now(they don't even put a mass simulator on the first flight tests, and won't put one on the next one) , well technically first of the version2 booster and starship can be builded and launched before the end of 2024, but we'll see.
@@theOrionsarms You're judging use once throw away test articles as if they should be mission ready. That's a flaw in your thinking.
@@TheEvilmooseofdoom that was my point, those pathfinder prototypes aren't real operational products, and for God seek stop disagreeing with people even when you are saying the same thing.
Well this aged poorly, we now know that Starships is massivly missing its performance target and it's present configuration (which is like the 5th iteration) can't even match New Gleen payload. And I can tell you why, because they didn't design the rocket first, they tried to design it iterativly like it was software. But you don't have a weight limit in software, you can 'patch' in code without having any effect on the rest of the system and when your nightly build crashes it dosn't destroy the server. Add to that the completly boneheaded move by Musk to use stainless steel, a choice made only to support rapid iteration but which clobbers performance forever, in software rapid prototyping is key, but it dosn't permently saddle you with low performance so again inappropriate software engineering logic doomed them.
I kindof feel you didn't actually watch the video...
When SpaceX did Falcon 9, they chose a conservative design that leaned on prior art - both for the vehicle and for the engines. They then upgraded the rocket and developed first stage reuse. They blew up a lot of first stages in that process because there was no prior art to lean on.
For Starship, they're in a similar situation - there is no prior art around second stage reuse. You assert that they should have designed the rocket first, but since nobody has done this before, it's not clear how they would know how to build a design that works. They flown the whole stack three times; I think the first flight was likely a mistake given the ground damage, but flight 2 and 3 have been pretty successful in rocket terms; flight three would have been termed a success if they weren't trying to develop a reusable system.
I don't understand why you would compare starship to new glenn. Starship is in active flight test, and all we have for New Glenn payload numbers are the assertions of Blue Origin, and we won't know if they are right until they are flying and recovering the booster.
Finally, I think you should go watch this video: ua-cam.com/video/NggG-XaVqpc/v-deo.html
Zero successful launches. Zero successful landings. No lunar lander. No tanker variant. No orbital fuel depot. 20 launches for 1 moon shot. Yeah huge success fanboys.
It's a system in development. A concept that clearly sails right over your head. Then again intelligence is NOT what you haters are known for.
Exactly. The argument in this video is Falcon 9 blew up so many times because there was no prior art for returning and landing a booster.
Ok, but there’s also no prior art for
- Super Heavy landing
- Starship fuel transfer
- Starship re-entry and landing
- Raptor relight in space
- Starship lunar landing
- Starship lunar ascent
So by Falcon 9 standards we should expect to see half a dozen explosions per evolution before we have a reliable system? And even then, we’ll probably blow up 3 or 4 more when other unforeseen shortcomings arise?