blakemw

The complexity with command pods is they have a low internal temperature tolerance, skin temperature is basically "instantaneous", if the skin reaches that temperature even for a moment the ship explodes. Internal heating is more insidious, because heat slowly bleeds into the interior from the skin and this can lead to the pod exploding even though its skin temperature was never exceeded, the critical factor for internal heating is duration of exposure and area of skin directly exposed. It is almost always a bad idea to have a command pod type thing directly exposed to the hot air, for Mun and Minmus return you can use random crap as a pseudo-heat shield, one example is keeping a 1.25m decoupler, flipped so it remains attached to the base. It has an internal tolerance of 2000 so it will protect the command pod. It may not matter if it burns off halfway through, because it still dramatically reduces the duration of heating for the command pod (and the command pod itself has like 2400 tolerances or something). So the answer differs based on whether the command pod is "naked" or has something up front to prevent internal heating. Naked pods tend to have to use two (or more) passes, exiting out the atmosphere to give them a chance to cool down. A protected pod can usually dive a bit deeper and do it in one pass. The altitude depends a lot on the draginess of the capsule, draggy capsules or those with drag enhancement (random excrements surface attached) can use an altitude of about 50km for a single pass, less draggy ones might need to use 35km or so.

The key to a wobble free station is clever use of autostruts. Autostruts are also the key to a station which violently tears itself apart. So getting a wobble free station involves mastering the arcane arts of autostrut mechanics. When a whole lot of autostruts converge on one part, this creates a concentration of forces which results in the wobble of doom. This will be what tends to happen with root part, and usually heaviest part and in some cases grandparent part, but not in all cases. The trick is how autostrut combines with symmetry placement. When autostrut targets a part with "symmetry siblings", a strut is created to ALL the symmetry siblings, who knows why, it is just how it works, this works whether the target is "acquired" via grandparent part or heaviest part (and with heaviest part, it doesn't matter if some of the syemmtry siblings are less heavy, the game first finds the heaviest part, which might be a full fuel tank, and then creates autostruts to it and all it's symmetry siblings, some of which might be empty). This creates a star-like pattern of autostruts which is very strong and wobble resistant. The basic idea then is to make the core of your station a symmetry structure, for example a cross. This can be braced with symmetry autostruts, and whats more when something docks to a symmetry docking port, you can use "grandparent part" on the docked vessel to connect to all the symmetry siblings of the core: Works well with Heaviest part too: Understand and abuse this game mechanic and you can have rock solid stations which are pretty resistant to wobble of doom. Note: In general I would advise bracing the "symmetrical core" with conventional struts. Conventional struts remember their original length and try to restore a vessel to its original VAB shape. Autostruts are re-created every time a vessel docks/undocks or the game is loaded and so autostruts don't remember what a vessel originally was shaped like and distortions can accumulate.

Not really in practise. Unless you're doing something really dumb you'd be better off investing the mass/cost of the dawn + power system into chemical or nuclear engines for the portion of the mission where TWR actually matters. Particularly if you take cost into account Dawns, Xenon and their power systems are very, very expensive and are easily out-performed on a cost basis by the LV-N for any remotely heavy payload. If taking only mass into consideration, Xenon is pretty mass-efficient, and technically if you want to get maximum range out of like, a 50t ship, you could do so with multiple Dawns, once the payload is heavy enough the weight of additional engines becomes trifling. But bear in mind that high ISP is better for getting large amounts of deltaV (for Dawn, 10000m/s+) than for getting small amounts of deltaV more efficiently, that's just how the rocket equation works. Once you're in LKO anywhere in the Kerbol system is within around 2000m/s from LKO which is very comfortably done with LF/Ox. Once leaving Kerbin, most destinations offer very little oberth effect relative to the insertion/ejection burn, or gravity assists (Jool system). Eve and Duna do offer meaningful oberth effect, but both are so close you don't need high ISP fuels and both offer atmospheres for aerobraking too.

I'd give it a C-, a barest passing grade, though I'm not interested in the mission builder so that feature does not affect the grade either positively or negatively. The new engines are poorly balanced, with the Wolfhound being much too cheap and the Mastodon being too weak to lift a Saturn V KSP-equivalent replica (you had one job Mastodon!) and the Kodiak being literally equal to or worse than the Reliant in every way (why not give it +10kN thrust or something). It has flaws like Structural tubes having internal attachments nodes but not shielding their contents from drag and bugs like Wolfhound nozzles being indestructible in certain variants and Service Modules being indecisive about whether they are shrouded/open or not. 1 grade off for not balancing parts harmoniously with existing parts & their replica roles, 1 grade off for bugs and flaws, and a - just because. It still gets a passing grade because the 1.8m parts, 5m parts and engine plates add something to the game and the additional launch sites and quick-switch between VAB and SPH are also welcome additions.

The Oberth effect has diminishing returns for larger burns. Say for instance that a ship has just reached escape velocity from Kerbin (3400m/s) and is at periapsis. An additional 50m/s, thanks to oberth effect, becomes 585m/s, 500m/s becomes 1910m/s, 5000m/s becomes 7680m/s and 50000m/s becomes 53300m/s. It can be seen that the Oberth effect is much more effective at increasing the potency of a small burn than a large one. If you're doing a low TWR burn then hopefully you're also using a high ISP engine (if not you must hate yourself, there's really nothing to be gained from having a TWR of less than 0.5 with LF/Ox engines because at that point the engine mass becomes negligible compared with the fuel tank dry mass). With the high ISP engines Dawn and LV-N it's really easy to pile on lots of deltaV, you actually do lose a fair bit from not fully exploiting the oberth effect, I'd estimate that it can easily amount to 500-1000m/s (i.e. that's how much more the burn ends up costing). But since an LV-N stage can trivially have 4000m/s more than a comparable LF/Ox stage, and a Dawn stage can easily have 10000m/s more these "losses" are not a big deal, just have 2000m/s more than you anticipate needing and it'll be fine.

While true in general this may not be true when extreme deltaV is involved. For example, if it takes 160m/s to get into orbit, but the craft has 5000m/s of dV, an extra 300m/s of gravity losses may be inconsequential compared with the total dV losses from the added weight of engine. Even with a TWR of 1.01, if you do a constant altitude burn ~15% of the thrust is available for building horizontal velocity. Dawns are especially special in this regard because they are horrifically heavy in relation to their thrust, being for example 12.5x heavier than an Ant. In general when using LF/Ox engines you want to aim for a TWR of around 2.0 to reduce gravity losses, but for Dawn and LV-N it's acceptable to go lower because their fuel is very light in relation to their weight (for example 0.25t of Xenon is nearly a third of the largest Xenon tank so if you're not wasting more than a third of a large xenon tank to gravity losses - a burn time of about 1 hour for 1 Dawn - you don't need to add another Dawn), but it might in some cases be even smarter to add a small LF/Ox stage to provide high quality thrust - for example an OscarB + Ant has the same wet mass (actually a little less) as a Dawn has dry mass, and the Ant would have a burn time suitable for like, ascent from Minmus - once the engine has burned out the dry mass is only 1/5th that of a Dawn so regardless of when the higher TWR is needed, a Dawn + Ant+OscarB, will have more deltaV than 2x Dawn. (alternatively, if the high TWR is needed very early in a mission, a Spark + FL-T200 has slightly less dry mass than the dry mass of a Dawn, so hauling around an empty Spark+FL-T200 has about the same impact on total dV as a second Dawn engine, but with a lot of high quality thrust early in the mission) In general my philosophy is the correct number of LV-N's or Dawns is almost always "1", and if higher TWR is needed at some point during a mission it's better to include a high thrust engine providing that much high-thrust deltaV - this works both for the 0.25t Dawn and the 3t LV-N, both engines are heavy enough in relation to their thrust that you can alternatively invest in a meaningful amount of LF/Ox.

That's not my experience. I've generally found that disposable rockets are around 20-30% more expensive than recoverable Twin-Boar SSTO. The cheapest an Orange tank has been put in orbit with a fully disposable launcher is just shy of $30000 (it was mostly kickbacks, with a Poodle core), whereas a Twin-Boar SSTO which cheats a little with SRBs to get moving can do it for around $20000 if the core is recovered at 90% (the cheating is because a full Orange Tank is too heavy for a single Twin-Boar to comfortably lift, a couple of Thumpers or 4 Hammers greatly reduce gravity losses early in the launch and dramatically improve the economics for this particular payload, a SSTO wont tend to be economical if the launchpad TWR is less than 1.4 because too much LF/Ox is wasted fighting gravity). Even if ONLY the Twin-Boar is recovered (everything else is shattered on impact with ground/water) a Twin-Boar SSTO can still be slightly cheaper than the best disposable rockets. It is important to get a high recovery %age though, if the booster is going to be dropped a quarter of the way around Kerbin you may as well just use Kickbacks.

One of the simplest balance changes to the Wolfhound would be increasing the cost. In career one thing which limits the usefulness of the LV-N is the 10000 funds price-tag, for instance that buys nearly 4 Kickback booster so if you're cost-optimizing you need to consider whether it's better to invest in a LV-N or larger lower stages. One problem with the Wolfhound is that it doesn't even have a high price tag. It's most readily compared with the Skipper or LV-N, yet it actually has a "thrust-cost-ratio" nearly twice as a high as the Skipper: Engine Thrust Cost Mass TWR TCR ISP Skipper 650 5300 3t 22.1 0.123 330 Wolfhound 375 1680 2.5t 15.3 0.223 412 LV-N 80 10000 3t 2.72 0.008 800 Since these 3 engines are all comparable in weight, it would be logical to expect the Wolfhound to be somewhere in between the Skipper and LV-N in all columns, if we were to pin the ISP at 412 and extrapolate the other columns, then the cost should be around 5800 funds, and the thrust around 340kN.

Because it is a primary launch engine, an engine designed to burn from sea level to orbit and its primary role is to lift stuff out of gravity wells, hence a jet mode and a decently thrusty rocket mode. We can compare it to very SSTO capable LF/Ox engines which are also suitable for launch-to-orbit use, the Twin-Boar has a vac ISP of 300 and the Mammoth/Vector has a vac ISP of 315. To me it wouldn't seem fair if the RAPIER had a better vac ISP than other launch-to-orbit engines. And a vac ISP of 305 isn't even bad in stock KSP, I'm perfectly happy using a Twin-Boar for an ejection burn (i.e. Falcon-Heavy style where the core gets to orbit with 1000m/s leftover). The rocket equation is not brutal at these ranges and factors other than ISP can overcome low ISP (for example, the Twin-Boar despite having below average ISP, produces thrust really cheaply and has really good TWR). There does come a point where ISP becomes of overwhelming importance, like Moho return or something, but for delivering stuff to LKO it's just not important to have a high ISP and that's what the primary launch engines are balanced around. If you want good vac range, slap on a dedicated vac engine. It is not valid to apply real world engine stats to in-game engines because the universe is different, of course when using an upsized game universe it is fair to increase the stats of engines to closer to real world values, but there is no need to single out the RAPIER in this regard as it is already one of the more highly powered engines for its niche, and if you want to be fair to the real world performance what you'd want to do is use realfuels/realengines style configs to make it actually use hydrogen with all the associated downsides of hydrogen.

I'm in the same camp, though I still use Mk2 parts for the utility (i.e. cargo bays), just never for dedicated fuel storage. I've found in testing that a single RAPIER can push about 20t of Mk2 plane through the sound barrier (in level flight across the ocean). Or about 30t of 1.25m/Mk1 plane. The payload penalty for Mk2 is very large, it can be reduced a bit by adding more wings allowing a better angle of attack (which really makes the lifting body aspect of mk2 a joke), but much of that added plane mass is wings, 1.25m/mk1 planes can go really light on wings. Mk3 parts perform more-or-less identically to mk1 parts, maybe a little worse. 1 RAPIER can push about 35t of 2.5m plane through the sound barrier, it is definitely more than mk1/mk3, but more of a 10-15% increase than the ~50% increase from going from Mk2 down to Mk1/Mk3. I find that Mk3 is usually worth it for the benefits they bring: improved impact and thermal tolerance, what seems to be better structural strength (could just be due to lower density, opposite to how Large Holding Tanks love to come apart at the joints), large flat surfaces for convenience of surface attachment and utility parts like LF fuselage, passenger cabins and cargo bays. When comparing with Mk2: smaller planes don't tend to have thermal or structural problems (it's the old square-cube law at work) and the Mk2 parts are a simply awful shape to work with, which leaves utility as the only reason to use them.

Well I was coming from the angle of it being a 2t jet engine (compared with 1.8t for Whiplash), the RAPIER and Whiplash when treated as pure jet engines are fairly comparable, the Whiplash is a tiny bit (almost but not quite unmeasurably) better at pushing planes through the sound barrier and uses a little less fuel, the RAPIER gets an extra ~150m/s top speed so the plane needs less LF/Ox for orbital insertion. Putting aside the cost, my preference for spaceplanes is the RAPIER even as a pure jet engine i.e. for Jet + LV-N, or even Jet + Aerospike. Given that a Spaceplane needs jet engines and the RAPIER is equivalent to a Whiplash you're essentially getting a rocket engine for free in terms of mass - or maybe 0.3t, but you don't even have to use the rocket mode for it to be a good jet engine. And if you're only going to LKO (which kind of makes sense with spaceplanes, not much point dragging an entire plane through the void) then you only need ~800m/s out of rocket engine and low ISP is really not a problem to LKO, high ISP only truly becomes a large benefit when going much further in terms of deltaV and when the rocket equation really gets brutal. But of course it's common to add a LV-N for this circumstance, the RAPIERs can still provide high TWR for lifting off from high-gee worlds (which whiplashes can't) while the LV-N provides long range. It can be argued that the RAPIER in rocket mode has more thrust than it needs (about twice as much in the way players often use RAPIERs) this means spaceplanes can work well with a 50/50 ratio of RAPIERs to Whiplashes, though that is only a cost-saving measure and it's only because of poor use of RAPIERs, when using them to the limits of their jet mode (i.e. a SPH TWR of 0.3) all the rocket mode thrust is useful. Even taking into consideration cost RAPIERs are a remarkably economic engine if you push them to their limits. A 2x RAPIER spaceplane can deliver about the same amount of payload as a Mainsail SSTO and at about the same upfront cost (there is huge scope for scrooginess when building planes...). In fact, if it weren't for the fact that spaceplanes are an order of magnitude harder to build than rockets, I'd say the RAPIER is actually overpowered rather than perfectly balanced, but since most players are using only a fraction of their capabilities it's probably fair.

Works great for me.

You can use autostruts (and actually, just normal struts) to stiffen up planes, the restoring forces provided by struts aren't vulnerable to physics warp in the same way as normal inter-part connections. Also if you're into mods, then Atmospheric Autopilot (the fly by wire function) results in rock-solid handling even at 4x physics warp.

There's a generally more scalable approach than tugs: Once you're in the end game and using Mammoth engines, a Mammoth based rocket can easily SSTO to Minmus, and even before then, cheating a little with some SRBs makes it pretty easy for smaller rockets (Rhino, Twin-Boar or Skipper scale) to basically SSTO to Minmus. The way to do this with a staged rocket, is you either empty out all the upper stage tanks or crossfeed them down into the Mammoth (for Xenon, just leave them full). Then you land on Minmus and get fueled up by an ISRU rover (which should dock using a Klaw). At this point configure crossfeed for staged usage as required. This approach can easily give an extra 3000-5000m/s of effective dV by re-using the launch stages. So once you have refueling at Minmus, the single-stage-to-minus /multi-stage-from-minmus is a very powerful technique, with the biggest benefit being that by refueling on the surface of Minmus huge amounts of fuel can be shamelessly expended. (I developed this technique when I got interested in using Triple-Mammoth from Minmus type vessels and the tugs and tankers were becoming problematically large).

I haven't able to mine asteroids in 1.4.3. I put it down to a mod messing with things but the situation continued even after I removed the mod. I thought it might be an "old asteroids" problem, but I didn't investigate the problem further and just gave up on asteroid mining.

I tend to favor Rocket SSTOs which cheat. Basically a big LF/Ox engine with a pile of fuel on top, then some quick-burning SRBs (Hammers or Thumpers) to get it off the pad. The core makes it into orbit and can be recovered. When you add small SRBs to a SSTO you get substantial increases in payload to orbit, for example a Twin-Boar can SSTO about 26t easily, with SRBs helping it off the pad it can do about 40t. The interesting thing is, that because SRBs are such a cheap way to reduce gravity losses, the cost per ton is about the same in both configurations, that is without the SRBs you have to burn a great deal of fuel getting moving with most that fuel being wasted fighting gravity, when you add SRBs, the SRBs are a bit more expensive than a recovered LF/Ox rocket, but because you can cheaply obtain a higher TWR much less fuel is being wasted fighting gravity.

I usually use 1 or 2, with 4x physics warp to speed things up. If I need more thrust than that I use an earlier stage with a LF/Ox engine to provide that deltaV. In particular nukes aren't fun for the ejection burn from Kerbin, the rocket equation still makes it easy to deliver LF/Ox to LKO, so just use LF/Ox. The further you've had to deliver fuel (in terms of deltaV), the more valuable a high ISP becomes. As a general rule, you want at least 15 minutes burn time out of a nuke - equivalent to 4 Mk1 Fuel Fuselages per nuke. I normally go with about 6 Mk1 Fuel Fuselages per nuke. If the burn time is shorter than 10 minutes you're probably better off just using an efficient LF/Ox engine. When playing Career, a LV-N costs 10000 which will buy nearly 4 Kickbacks, so the cost of a LV-N will fund a significant amount of lifting power, if you look into it, then a multiple-LV-N solution is often beat by removing all the LV-Ns except one, and using those funds to upgrade the lower stages.

I'm not that bothered by the 2 man can, probably because I use it for a control module for 2.5m stations and bases. RAPIER is perfectly balanced. It's pretty great as a jet engine, and not terrible as a rocket engine. The ISP is low for in vacuum because you don't need to carry additional weight in rocket engines (or jet engines), all things being equal if you shave off 2t of mass, you can bring 2t more fuel or payload so for missions to LKO it's hard to beat RAPIER spaceplanes. For longer range missions you might invest in higher ISP vac engines, but that's what balance is - an engine being good in some circumstances but not others. (also realistically, it doesn't have high expansion nozzles, so should have poor vac ISP) Mk2 parts could definitely use some drag reduction. Their low mass-performance compared with dedicated fuel tanks and dedicated wings (carrying less fuel, producing less lift and more drag) is only partially offset by their thermal and impact tolerance. In KSP thermal tolerance is often not a big deal, it is by no means impossible to re-enter an aircraft using the aircraft wings with 1200/1200 tolerance. However I find that Mk3 parts are fine, while they are a little inferior to rocketry parts they aren't bad enough that you really feel it, and the greatly increased structural strength of Mk3 planes is worth something - often my rocket-part planes (using parts larger than 1.25m) break apart if a mouse farts whereas I never have that problem with Mk3 planes. Aerospike is fine, Skipper is fine, Fairings are fine. Not sure about Mk3 cargo bay, I only make passenger and fuel tanker spaceplanes and the only reason I'd ever use a cargo bay is to house ISRU equipment and in that case the mass is kind of negligible. Mainsail could do with a small buff to help it compete with the Twin-Boar as a launch engine. Presently the Twin-Boar (deducting the integrated tankage) is cheaper than the Mainsail, has 33% more thrust, weighs only 0.5t more and produces more drag though that's pretty negligible for a 2.5m rocket. I'd give the Mainsail an additional 5 ISP, giving it a slightly more convincing fuel-efficiency advantage. MH Balance: Wolfhound ISP, mas and thrust are too high - I don't think it's ISP should be higher than 365. And Poodle should be depreciated (like the Mk1-2 command pod, not deleted, just not shown anymore). Skiff is fine. It has roles to play without being overpowered. Kodiak need something to differentiate it from the Reliant other than being costlier and draggier with a weaker alternator. How about 20 more thrust or something? It's just sad if an engine is (slightly) inferior in every way to another engine. Bobcat feels a little weak but I'm not convinced it needs a buff either. Mastadon badly needs a cost reduction so its usable in career. Matching node sizes just isn't justification enough in career to use an engine and I'd like it better if it's "thrust-cost-ratio" was on par with the Twin-Boar (which would be fair, for a garbage ISP engine). It could also use a thrust increase, it's okay if the ISP is bad as long as it's good at lifting heavy rockets off the ground without the help of SRBs.

With the modern aerodynamic model throttling is never efficient (except to reduce overheating) unless your rocket is super absurdly draggy which would be a bad design. With one exception that has nothing to do with aerodynamics: If you have a "falcon heavy" style rocket or just generally a core with side boosters design, it actually can be beneficial to throttle back just the core if the boosters are providing plenty of thrust (i.e. a TWR at least of 2.5). The reason this can be beneficial is that it saves fuel in the core to burn *after* the side boosters have been decoupled, that fuel then goes only to accelerating the mass of the core and not also the mass of the side boosters. The fuel will also be burned at a higher altitude at better ISP, which is relevant to some engines, but you'd do it even if launching from a vacuum world. This kind of core throttling is and has been done in real life and the reason is as much getting more deltaV as reducing pressure on the rocket, i.e. from the wiki for Delta Heavy rocket:

You might want to make them pointy, nothing is less draggy than a pointy fairing. They also aren't heavy, for example, a mk1 landing can enclosed in a pointy 1.25m fairing is lighter than a mk1 command pod. Because of square-cube law the larger ones are even lighter by volume enclosed, fairing walls don't get thicker with larger fairings so a mk3 fairing can end up weighing ridiculously little for the volume enclosed.

My rule of thumb is: use a fairing. Only real exception is for early career when there are no fairings, and for fuel tankers which can be made entirely from streamlined adapter pieces. Most parts which aren't inline fuel tanks or aircraft parts - things like science lab, hitchhiker can, etc generate quite a lot of drag even if they look like they fit the stack. Also there are some gotchas around surface attachment which can cause surface attached parts (solar panels, parachutes etc) to generate a lot more drag than you'd expect, you can learn all the gotchas and work around them.... or you can just use a fairing. Fairings also let you use vacuum optimized parts which are usually lighter than atmospheric parts and generate more drag during reentry making reentry safer (which is somewhat paradoxical, particularly as the part description usually warns you not to use these parts for reentry, but in KSP there is neither enough heating nor enough pressure to threaten vacuum optimized parts under normal reentry conditions). I've played hard career a lot and generally found fairings a good investment in that context, they reduce the possibility of things going wrong and improve the consistency of launches, so even when paying for the things I reckon they're good value.

I had a look, here's an aerodynamic overlay and the debug option to display drag values in action menus: I probably didn't find everything secreted away inside the vessel, but I found a bunch of stuff. The major cause of drag is the cargo bay nodes are wrongly connected, this is causing roughly 2 RAPIERs worth of drag in the case of the Mk2 LF fuselage, and 1 RAPIER worth of drag in the case of the MK2 RF fuselage. The rule of thumb is to never attach stuff which is outside the cargo bay to the inside nodes of a cargo bay. Attach fuselage only to the appropriate outside nodes. Also the deployable solar panels are not "inside enough" the cargo bay for the game to consider them shielded from drag. The two of them are causing about half a RAPIER worth of drag (those things in general are draggy pieces of excrements, even when placed edge-on to the wind which is the ideal case, use Gigantors pointy-on to the wind if you can). The Monoprop tanks are also contributing about half a RAPIER worth of drag, I don't think they are even trying to be in the cargo bay. They need to be in order to be shielded from drag. Anything you want to be shielded from drag needs to be convincingly inside a cargo bay, the game errs on the side of making things drag unless it's absolutely sure it's inside a cargo bay. Fix those issues and you would eliminate about 4 RAPIERs worth of drag, and could remove that many jet engines and still have enough power to get through the sound barrier. (but I'm by no means confident I found everything "sneakily" generating lots of drag, in particular, I'm sure there's something dodgy going on with the LV-N attachment) 2 RAPIERs can push a fairly sizable spaceplane through the sound barrier in level flight across the ocean, at which point the mach multipler kicks in with a vengeance allowing the plane to do an accelerating ascent. One of my better spaceplanes is my "Duna Operations Plane", it's around the limits of mass & drag which 2 RAPIERs can get through the sound barrier: So your plane should be easily able to get by with 2 RAPIERs, if all the unnecessary drag is eliminated.

To a large extent this is a matter of poor engine to propellant rations. My rule of thumb: each RAPIER should have about 3/4 of a Mk1 Fuel Fuselage and 2-3 FL-T800's worth of LF/Ox each whiplash should have about 2/3rds of a Mk1 Fuel Fuselage each LV-N should have about 4-6 Mk1 Fuel Fuselages Using these engine to propellant ratios give a very respectable payload to orbit, or to further destinations. Turning it into a plane that can get into orbit is a matter of wings and streamlining. A good acid test is the plane should be able to break the sound barrier (get up to 500m/s) while flying level across the ocean at low altitude (<1000m), if it can't then you can try: reduce drag by improving streamlining (aerodynamic overlay helps), add wings to increase lift (allows a shallower angle of attack which reduces body drag), reduce weight, or add more jet engines (with an appropriate amount of fuel).

Grandparent is much safer because it's much more limited in its scope, but it's also less powerful - except when dealing with symmetrical placement, for some reason autostrut connects to the grandparent and ALL the grandparents "symmetry brothers", a simple example: This is useful for creating large, rigid, stable structures. Importantly, unlike "root/heaviest part", which concentrates forces on one part, symmetry grandparent distributes forces. It's an incredibly powerful technique when used deliberately and it's almost always a good idea to try and make the core of a station a symmetry structure, as an example: Because the station was built using symmetry, now when the Jumbo-64 fuel tank is set to "autostrut: grandparent part", it generates 4 autostruts, widely dispersing the forces over the station.

Note that ALL the issues from autostruts CAN be replicated by normal struts, it's just that autostruts make it really easy thanks to limitless range and symmetry. But when you have heavy stuff on wobbly arms struted to other stuff, that is when the wobble-of-death can happening. Note that (perhaps surprisingly) moar struts helps with this because suppressing the wobbling prevents the problem, but it also means there is more pent-up energy in the station which can cause other kinds of kraken attacks, no struts has a tendency to be most safe against kraken attacks, with lots of struts being better than few struts. (unless those few struts are carefully placed - it mainly is heavy weights on wobbly arms that cause the wobble-of-death, so for example autostruts running the length of the vessel from the probe core to the engine can be safe because there is no leverage).