FleshJeb

I have a lot of meta-questions about the process of game QA and the state of game development in general, and this is a golden opportunity. However, I don't have the energy and it's frankly a trivial subject. So, I'd just like to encourage everyone to donate to LGG's Patreon, because his work essentially carried most of the KSP franchise on his back for years, and the number of Patrons he has is shamefully low. I'm once again happy to say that I paid him more money than Squad ever got out of me, and he deserved every penny. I haven't played KSP in 3 years, and I'd still be happily paying him if it were in the budget. A very honorable mention to HebaruSan as well. Thank you both for your tireless efforts in making this absolute disaster of a franchise worth playing.

Yes it is, person speaking from a position of authority... https://bugs.kerbalspaceprogram.com/issues/27484 https://bugs.kerbalspaceprogram.com/issues/13366 Still a problem as of 1.6.1 and 1.11.2. per the bug-tracker. EDIT: Still true for cargo bays as of May 6, 2021. As far as I know they use very similar mechanics and are subject to the same bugs. (Note the carefully qualified answer):

I haven't had KSP installed for about 3 years, but I can tell you that fairings and cargo bays, etc. don't occlude drag when they're the root part. It's a bug, but you should be able to re-root the craft (4 key) and it should fix it. If not, try re-rooting, pulling the fairing base off, and installing a new one. I think that's the likely solution, but a picture would make it easier to eliminate other causes.

Cautious and deliberate engineering that meets or exceeds all its test objectives is SOOO last-century.

If anyone wants to watch the signal disappear, it's at 16:45:08 UTC in Steve's livestream above. (Clock in the bottom left, I don't think I can link to a timestamp.)

Site evaluation, design and calcs for the retrofit, prepping the site, formwork, and then pouring concrete. After the cure they can backfill, compact and start working on the slab. All those steps take time, and you can only fit so many men and machines in that hole. They might only be able to get two excavators working concurrently, and they have thousands of cubic yards of material to move. Remember they have to make the hole bigger and get it controlled before they can make it smaller again. I've worked on a few landslide repairs, and they have to pull out ALL the loose stuff and then some. This is all assuming that everyone and everything they need is on-call and working overtime.

I'm actually presuming the piers have most of their original strength. The above-ground portions were steel-jacketed--They're fine. I have to revise my earlier estimate and say this is the equivalent of a seismic retrofit. They can jacket the existing piers with more concrete and rebar, and add some shear structures between them if necessary. Any post-tensioning cables can go outside all that. So we're down to 2-3 months and $5M.

We're living in the future. This is really cool!

Disclosure yet again, I'm a civil engineering/land surveying technician without a degree and licensure--I'm making some minimally-educated guesses here. I happen to have some rudimentary understanding of adjacent disciplines because I have to communicate with them and produce work that interfaces with theirs. I also have to collect much of the data they need before they can start, and then translate their finished product back on to the ground to be built. No recollection, I didn't design it, I just studied the structural plans so I could show them where to build it per the design. Structural engineering above a certain scale is usually dominated by "live load" rather than "dead load". This means wind, water, or seismic forces are going to set the minimums for the design. (On this one, the water forces vastly exceeded everything else--which is why it was $5M to build rather than $500K.) It's also why bridges and buildings can sustain significant structural damage and remain serviceable for a long time when not experiencing extreme conditions. I'm not even going to speculate on what magnitudes of loads that rocket blast imparts, but I bet it's fun! The only limiting factor in engineering is money. My bridge project took two years of intermittent construction work, because public infrastructure is typically very cost-focused. People are making a lot of jokes about free digging, but the site cleanup and prep alone could easily be a month. See my diagram below, but I think those big horizontal tubes are post-tensioning ducts, and the cables in them are shot. There might be easy access to the end anchors, or they might have to jackhammer them out and find a way to bond new ones in to the structure. If the piles or caps took flame damage, they'll have to support the structure on some falsework, cut out the damaged portion, and find a way to splice new rebar and concrete in. One of the things that confuses me about this image is that while there's a fair amount of steel on the ground, there should be a web of rebar thick enough to walk on in that hole. It can't all have vaporized or blown into the ocean, so maybe they went really cheap on the reinforcement. Absolute bare minimum time with on-call engineers and construction crew might be four months. Tying rebar cages takes time, even if you prefab them offsite, and concrete takes 28 days of curing to meet design strength, and they can't do this all in one pour. Six months would still be a rush job. (Of note, I'm renowned for underestimating time, so you could double these numbers and not be unreasonable.) EDIT: This is in the neighborhood of a $10M repair, I'd be willing to believe $15M-$20M of effort for a rush job and premium prices on materials. Anything much less than that is a lie or involves chattel slavery. Just by way of illustration, this is what the site might look like partway through the repair process. All those steel beams and wood are temporary, and take a long time to set up (and are done in a fairly precise manner).

Down to local bedrock is the answer. I construction-staked a bridge a couple years ago that had similarly-sized pillars (5 ft dia, instead of 6 ft) about 75 ft long, extending about 50 ft below ground level to bedrock (This bridge was hugely overbuilt because it's designed to survive being under 20 ft of moving floodwater). It's impossible to tell from any of those photos because they're splicing the rebar into the already poured below-ground portions. 100 ft deep wouldn't be shocking, but neither would 50.

I'm this kind of KSP player: Don't pee on my head and tell me it's raining--I've been in the civil engineering industry for 30 years, I have close to 1000 projects under my belt, and I'm the guy they call to fix it when the project is FUBAR. What, in your mind, would have constituted a failure for this mission? Not clearing the pad? They're very lucky it made it, for completely predictable and avoidable reasons. Luck is not in the engineering handbook. This is a false dichotomy. The industry can be less risk averse AND dot their i's and cross their t's. That doesn't change the fact that with some better risk mitigation, this mission could have had a MUCH higher return on investment. That's objective fact. My subjective opinion is that I find it hard to believe that the value they got out of it was worth the investment, since an orbit-capable Starship got all blowed up before they lit the candle. I want to see humanity have access to space as much as you do, that's why i think it's important to recognize and critique the failures. Yes, I 100% agree. The Shuttle was a waste and a boondoggle from first engineering principles, and I have nothing nice to say about it.

I have some issues with the narrative that this was a wild success. Smells like marketing. Shifting the goalposts from aspirationally "crashing Starship in the Pacific" to "clearing the pad" is a pretty clear sign to me. "Flying concrete is bad for rocket engines" is not novel or valuable data. "Well they figured out how NOT to build a pad!" Yeah, NASA figured that out over 50 years ago, and they have the papers and math freely available to study. This is still not a lesson that needed to be learned. It's a big rocket, but the parameters are not out of spec for what we already know how to do. I sat down with my dad this afternoon, who has been a civil engineer for 45 years, and described the setup and end result to him. The first words out of his mouth were, "That is some amateur BS." He asked me why they launched if they had a flame diverter almost ready to go, and I was only able to answer with "scheduling issues." His response was, "Of course schedule is always an integral factor, but if your 'profit margin' is [volume and quality of] test data, and you lose half to 3/4 of your opportunity to generate data [Starship not getting any meaningful testing in this event], then you've just cost yourself time and money for no good reason." In terms of opportunity cost, I can only see this as a massive failure. I will grant that erosion is a big problem in engineering, because it's a cascading failure--Any flaw in design or construction will be exploited. So the answer is always robust design and safety factors. At the very least, you should understand the failure modes and plan for where the blowout is going to occur. If anyone followed the Oroville Dam spillway failure, you might recognize that it was located well away from the main body of the dam, and that's why the undercutting wasn't completely catastrophic. It was still a massive design and maintenance screwup, but nobody died. Sevenperforce did a really great analysis upthread showing that being only 18% above nominal in a bunch a factors leads to a 600% worse outcome. In isolation, that seems like a big number, but it also means that a very tiny 1.2 factor of safety on all their inputs would have solved the problem. I've worked for engineering firms that specialize in doing challenging and innovative things. Unfortunately, that didn't include having robust standards, workflows, and competency in the basics. Because we're cool and innovative and that's just boring stuff. What that results in is failures, blown budgets, late nights, apologetic calls to clients, and construction change orders. The cool and innovative thing was rarely the primary challenge on the project. So when I see an organization making very basic engineering errors and trying to handwave them by pointing at the cool and innovative part, I'm not too impressed. I just see a cultural and systemic problem in the organization.

It's somewhat comforting to know that all young men, regardless of location, are subject to the same brands of stupidity. The essential problem with hero fantasies is that they require someone else to get hurt first. Old man wisdom says you should spend your efforts preventing the problem from ever arising. Boring, I know.

"I could log into a dressage forum and start asking for recipes."

You neglected, "Off a fellow mammal that's recently dismounted from a brass pole."

This is why I'm taking the short odds on that tank farm next to Starship being an extraordinarily bad siting decision.

As a connoisseur of weird porta-potty graffiti, I support this measure.

Do you think that there might be value in this, he asks Socratically?

Dearest Sir, I would like to complain that I did not receive my participation trophy for nominating the KAL thread. I believe I did it very early in March, and followed the nomination instructions explicitly. This oversight shall not stand! --FleshJeb I'm kidding of course. Thanks for doing these and I hope you're as well as can be expected.

It's the NorthySouthy Method. UppyDowny would be and

The one thing left out of Snark's guide is to point the docking port of the target ship to Normal or Anti-Normal . This simplifies the complexity of the operation a lot. If it's pointed in any other direction, the burns you do to get from 100m to 0m will have a noticeable effect on the size of your orbit, and you'll have to correct for them. With this trick, mostly what changes is your inclination, and you can almost ignore the other motions. You can actually dock fairly easily without setting a target, just by locking your orientation to the opposite of the target, and you'll remain parallel to it during the entire process. So it just becomes a 3D translation problem with no rotations to worry about.

See, what you have to do is conduct a lot of research on a microfiche; then you swear never to voluntarily look at light text on a dark background again.

I played 6000 hours of sandbox KSP 1, I never get bored, because I have an imagination. EDIT: LOL, Joe, I didn't even realize this was your thread. I'm NOT picking on you, I just don't understand this "boredom" thing.

I may have spoken rashly. Welp, at least I'm on-brand.