Hejnfelt
Members-
Posts
114 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Everything posted by Hejnfelt
-
If you look in your 1st post in this thread you have a list of images for parts. There is a picture of a part called "Winch port M" and this part is missing in the editor. All the tutorials and reviews of KAS specifically mention how important this part is and without it, the mod won't work. I am trying to build a skycrane with your mod and when I attached the winch to a grapple or magnet without this part, it seems like the winch isn't actually connected to these parts at all :/
-
Hello I am missing the "winch port M" and cannot get the winch to work without it. I've tried installing with CKAN and Manually but part is still missing. I also have KIS installed. Am I doing something wrong? I have reproduced this issue in a clean install with only KAS and KIS mods installed. The little winch port connector which is crucial for the mod simply doesn't exist anywhere in the editor.
-
Mod-Oriented Tech Tree (Aug 3, v0.3.3 bug fixes)
Hejnfelt replied to cvod's topic in KSP1 Mod Development
I just found your tree and am very impressed. I recently restarted a campaign with ca. 50 mods and am playing with the near future technologies tree. I decided not to install Interstellar due to using the NF tree. My question is whether I can install Interstellar and then use your tree and nothing will be broken? Or whether I will have to start over I realize it's impossible for you to guarantee nothing being broken, I guess what I'm after is sort of an educated guess whether installing Interstellar this late in my career and using your tree will break my save or not. -
Perhaps update Firespitter?
-
Haha, maybe my screen isn't correctly calibrated
-
I created the SAS and RCS parts I was talking about. Here's a showcase video that also includes ObsessedWithKSP's decoupler because I thought that was pretty nifty. Parts copied from Porkjet's dronecore Parts color-coded somewhat according to KSP principles. RCS is yellow, decoupler is red and SAS is orange Get all the parts here - Download original look from Mega - Download colorful look from Mega This redistribution has been permitted by Porkjet.
-
This is career mode and the Rapier is unlocked very late and so is the double engine adapter. Clipping means risk of bugs. I'm finally playing a career playthrough which I expect to finish and am not willing to run the risk of losing Kerbals due to clipping. While I appreciate all the suggestions, these are simple things that I've ruled out for design reasons. I like pretty planes The Dronecore adaption fits my purpose.
-
Presenting the Kerbin Space Taxi. Fully capable SSTO. Great for those "save a Kerbal" missions. It uses the RCS engine to circularize and regular RCS thrusters to dock.
-
That's actually a great idea. I didn't even see the DroneCore because it also is unlocked very late in career mode. I will look into doing this little project later tonight.
-
Yes, but in career mode, the cargo bay is unlocked very late and thus you're stuck with no SAS but what's in the cockpit.
-
This pack is great, if still missing some essentials. I spent half an hour today refitting B9's attachment nodes to fit flush with these parts only to find that B9's Mk2 parts aren't actually the exact same size as SpacePlane+ parts. The parts I miss in this pack are mainly: - Mk2 SAS reaction wheel - Mk2 RCS tank Any chance you'll work on those?
-
[1.10.0] Final Frontier - kerbal individual merits 1.10.0-3485
Hejnfelt replied to Nereid's topic in KSP1 Mod Releases
This weekend I restarted career mode with over 50 mods of which I'd never tried most of them before. I can happily say that this is in my top 5 favorites! I highly suggest it's implemented in the final stock version. It's non-intrusive and rather than just having a "favorite" Kerbal, I now see who's most distinguished. I don't just use Jebediah only anymore. Bill and Bob now have their personal specialties as noticed by their individual merits. Jeb is my go to rocket pilot and Bill is becoming my go to spaceplane pilot. Bob stays with Kerbin SOI and fly's mostly Base missions between Kerbin, Mun and Minmus. I've started actively recruiting Kerbals for specialization in certain conditions such as base crew and shuttle missions. There's a new feeling of "rank" within my hired group of Kerbals as some have achieved certain things and others have not. This has led to establishment of small requirements where I won't use Kerbals without cetain experience for certain missions. Thus I fly "training missions" where skilled pilots "train" young recruits to a certain extend. For example I have a tiny agile plane in which Bill takes out new recruits and subject them to 10-15G in turns to get them their G force badges before they get to go in a rocket bound for space. I also have trial lander sessions in which Bob trains recruits in taking off from the launch pad and landing in the grass close to the KSC. Seriously, this mod has one of the largest impacts on my game play due to how well it's fleshed out and integrated into the Kerbal universe. If I should add one request it would be the badge notification window post flight is integrated into the stock tabulated science and funds info overlay. Thank you so much -
[0.90.0] Fine Print vSTOCK'D - BETA RELEASE!!! (December 15)
Hejnfelt replied to Arsonide's topic in KSP1 Mod Releases
Hmm it seems something is still amiss. The old contract I had still exists in the "active" contracts board. That particular orbit doesn't show in the tracking station anymore though so now I guess I still have to "decline" that contract and take the loss to get rid of it. I do however have 2 other orbits in the tracking station for 2 new contracts on offer. -
[0.90.0] Fine Print vSTOCK'D - BETA RELEASE!!! (December 15)
Hejnfelt replied to Arsonide's topic in KSP1 Mod Releases
Thanks a lot for the clarification. This mod is one of the most awesome I've tried yet. It really improves gameplay. -
[0.90.0] Fine Print vSTOCK'D - BETA RELEASE!!! (December 15)
Hejnfelt replied to Arsonide's topic in KSP1 Mod Releases
Great thanks. So if I patch while I have the contract with satellite orbit apoapsis outside Kerbin's SOI, will it disappear without me having to pay the failure fine? Also I second the question of whether stock contracts will disappear? -
[0.90.0] Fine Print vSTOCK'D - BETA RELEASE!!! (December 15)
Hejnfelt replied to Arsonide's topic in KSP1 Mod Releases
The satellite deployment missions are broken. I accepted a mission without being able to see the numbers. When I got the satellite up and tried to get it into the shown orbit, I realized the orbit is outside Kerbin's SOI!!! This is a contract that would have supposedly net me a nice sum, but now I have to take the loss on the contract because I can't create the orbit due to escaping Kerbin. Had I known the Pe and Ap of the orbit was over 84Mm I would have not taken the contract. Now I have 3 more satellite deployment missions but I don't want to accept any of them because I cannot see which of the missions are which. One of them I want to do but I can't tell them apart. Please include the actual Pe and Ap orbit numbers in the contract guidelines. -
WASP SSTO 2.1 tonnes Around Kerbin in 41 minutes with only 45 fuel and 22 oxidizer. Does it matter I used B9 landing gear? www.youtube.com/watch?v=gG3B-0DyLWo
- 3,147 replies
-
- spaceplane
- k-prize
-
(and 2 more)
Tagged with:
-
10 char
-
Tavert! I do appreciate it. I even tried tipping you money (dogecoins) on Reddit for this, but you never accepted, so the transaction failed I'm glad you appreciate it. I think Mach 5 may be possible but I have not come across a design that can achieve it. Yet... As seen in my video, they don't really cover anything. The weird part you can see when I ditch the fairings is just the fairing adapter. The fairings reduce drag quite tremendously and with FAR, nosecones actually benefit you and thus with fairings you can build very specific nosecones as seen in my latest video. Whether the conic fairings are better than the eggshaped fairings is still up for debate.
-
It's not and yes the joints are becoming an increasing problem. So is impact force with the air if you can't keep your prograde vector and aircraft nose direction aligned. The F119 is most efficient and seems good, but the TurboJet is king above 1000 m/s. The F119 won't even reach 1200 m/s as it cuts off before then. 41800 m/s seems really really fast. Like >mach 100 fast. I experience thrust velocity curves that follow the cfg files to the letter.
-
Well it's a great challenge and anyone trying it will definitely learn something about aircraft design. I know for a fact, faster is possible. The Turbojet's velocity curve looks like this The thrust multiplier decreases very slowly in a non-linear fashion from max 1.0X at around 1000 m/s to ca. 0.5X at 2000 m/s. Meaning at 1400 m/s I was still enjoying around 80% of max thrust and definitely had not hit the limit on thrust needed to overcome drag. With weight optimization (fuel mostly) one can go even faster. But most importantly! I didn't make full use of my air on this design. Theoretically the radial intake should keep even more TurboJets running so perhaps someone will come up with an even better design. But then... You also have to fly it. And that's easier said than done
-
1,402 m/s - MACH 4.134 - B9 and Procedural fairings with FAR Equivalent of 5,047 km/h or 3,136 mph Well... First the proof: Top speed @ 1:05 Landing @ 10:30 Now for some comments. Flying the craft You may find it odd that I chose to fly the aircraft from the angle shown in the video. This is however fully conscious as flying it from a chase perspective gave me very little visual input regarding the position of the plane - i.e. the direction of thrust - in relation to its prograde vector. This is crucial since deviating outside or even close to the edges of the prograde vector on the nav-ball at >mach 3 speeds, will end in catastrophic failure. Observing the plane from a backwards facing fisheye perspective allowed me to use the conic fairings as a very precise indicator of the planes positional direction in relation to the prograde vector in combination with the navball. Craft design Initial design credit goes to user Sevant. This craft design is very suited for fast flights. Engines More engines means more thrust. But it also requires more air. I tried several designs featuring 2 radial intakes on the sides of the craft, but they generated so much drag that achieving +1200 m/s was a challenge. The centered radial intake in the front took many tries to perfect. Just a few steps up or down in the SPH and the intake would generate drag either pulling the plane towards the ground or towards the sky, making it impossible to fly above 500 m/s. There has been talk about engines in this thread and the fact is that the stock TurboJet is the best for > 1000 m/s flight. The B9 engines all cut out between 1100 - 1430 m/s meaning at 1000 m/s they are at 1 thrust multiplier and at 1430 they are at 0. The stock Turbojets however 0 out at 2400 m/s and achieve multiplier 1 at 1000 m/s. This means that when the B9 engines cut out, the TurboJet still enjoys around 75% of its thrust capacity. If you wanna go fast you must use the stock TurboJet. Anyone who wants to know more should check out the followign imgur album by user Tarvert. http://imgur.com/a/hyuPE#0 Tarvert is a source of amazing information and if you like looking at graphs like the one for the atmospheric engines, you will enjoy some of his other work submitted to Reddit: http://www.reddit.com/r/KerbalSpaceProgram/search?q=tavert&restrict_sr=on Stability Stability is another issue. As explained in my flight note, simply deviating outside the green prograde vector on the navball at speed will end in complete disintegration due to drag forces applied to the craft parts. Some parts even explode due to the impact with the air. My design features control surfaces for the purpose of stabilizing the craft at speed. These also help move the center of lift back slightly which helps tremendously in landing the craft and flying it at slower speeds. Stability is even more of an issue on the return flight as this design is very rear heavy. This is also why I drop the engines - or at least tried to with the reduced weight, it is easier to fly and land. You may think putting rudder(s) on your craft will achieve stability and you'd be right, but it puts unnecessary drag on your craft and your top speed will likely be < 1300 m/s. Craft shape I have tried long thin and circular designs and none of them match up to the wide and flat design initially shown by Sevant. I have also experimented with different fairing shapes and while the length of the procedural fairings do not seem to matter for speed, the longer fairings can help move the center of mass forward. The conic fairings work best for visual clues about the direction angle of the plane when looking back on it during flight. Landing You want to switch off braking on your front landing gear. If you find yourself landing at 100 m/s, braking on the front wheel can - more like will - be catastrophic. Also make sure your rear landing gear is somewhat spread out. With a think shaped plane you'll want to place your landing gear farther out on the tip of the delta wings than I have. Due to the wide footprint of my craft I put them closer to the body of the craft to decrease drag in flight and decrease flex upon landing or the belly of the craft may have impacted the ground. Conclusion and suggestions for other competitors Use TurboJets. Consider fuel necessary. Make sure you have some control surfaces for stability at max speed. Use the COM and COL indicators in the SPH during construction. Be extra careful during slowdown or this happens Failure @ 1:10
-
I'd like to point out that slowing down from above mach 3.5 below 1000km is becoming even more of a challenge than actually achieving the speed. At present moment I have designs which can reach mach 4 but cannot slow down due to instability from sudden lack of thrust and resulting complete disintegration Allowing people to climb above the 1km ceiling, greatly reduces the complexity of this challenge as all you have to do is climb until either your engines lose thrust or it is safe to slow down in thinner air. Then head back to the KSC. Half of this challenge is simply making it back to KSC and landing safely as is aptly pointed out in Sevant's crash videos. My suggestion is to keep enforcing the hard ceiling as it has been so far.