• Content Count

  • Joined

  • Last visited

Community Reputation

32 Excellent

About blakemw

  • Rank
    Junior Member
  1. You can always have a second tier of transformers. Solar Panels w/ Jumbo Batteries -> Transformer -> Heavi-watt backbone with generators and smart batteries -> Transformers -> Distribution.
  2. The simple way is to have the Slugs/Solar Panels directly charge Jumbo Batteries, that dump their power onto the main grid via Transformers. The other generators with their Smart Batteries should be downstream of the Transformers.
  3. In my games I have had 4 Plug Slugs. These each generate on average 50 watts when wild. So the 4 of them, produce 200 W on average (with a sufficient battery bank), the same as running on a Hamster Wheel for half a cycle. This proved plenty enough power for research and a few other things, for example the Research Station uses 60 W and the Supercomputer uses 120 W. In fact, after research, you should still have about half that power for other things (assuming only a single researcher). At the moment Solar is probably OP so makes a logical thing to beeline.
  4. I think there's definitely a problem with the solar panel ease. A few possibilities. The first would be to remove the free glass. The second would be to add a second, harder to acquire, ingredient to solar panels, such as steel or plastic. The third would simply be to reduce the sunlight intensity on the starter world so solar panels only produce like 100 W. I'm a bit torn on Slugs. If I'm not mistaken you have 4 wild slugs (at least both games I've started had 2 pairs of slugs). That's not very much power, but it is enough to get you to Solar Panels. And it makes the Swamp start a pleasant experience by more or less eliminating the requirement to operate a Hamster Wheel: in my second playthrough I simply didn't build a hamster wheel and it went very well, I had occasional blackouts late in the day but I felt it better than having some idiot run off and waste time on the hamster wheel (given that without Smart battery the dupe will run until the batteries are fully charged, which the slugs will do anyway). The real question of Slugs is how abusable they are via starvation ranching. FWIW, I despise the starvation ranching game mechanic, there should be penalties for malnourished and starving critters, or at very least, for being unhappy, since generally starvation ranching relies on glumness to work well (massively reduced metabolism, for no reduction in production), an 80% penalty for being glum would sure hamper things.
  5. I agree that oxygen masks should hold less. Maybe even just 10 kg (enough for 100 seconds of work).
  6. That's positive. I really, really hope that starvation ranching is going to go away, with some rebalancing of critter food requirements.
  7. Making builds that don't use Steel (or Thermium) is something I've been playing around with for about a year, but it took a while to come up with a design that replaced the Shipping components with steel-less substitutes that I was really happy with. The basic issues is that Iron machines have an overheat temperature of 125 C, but a Steam Turbine doesn't even start up until temperatures reach 125 C, hence a temperature differential "exploit" is required. The heart of this build is an Autosweeper in a Liquid Lock to the bottom-left of the Volcano. From this very particular position the Autosweeper can reach the tiles under the Volcano where the refined metal debris fall. The liquid lock also makes it possible to have a vacuum to the left hand side, and by positioning and orientating the Conveyor Loader correctly the refined metal is being stored in vacuum while the Loader is still being cooled. There are also a number of places where you could build Storages if you want to expand the vacuum storage of hot metal for slow release during the dormant period. Now I'm sure some players are already getting ready to type: "just have the sweeper grab through a diagonal", but good luck coming up with a diagonal sweep build as miserly in terms of components and as compact as this one. Now for how this design stays cool. The exhaust water from the Steam Turbines should be dripped directly onto the Autosweeper, this means the Autosweeper will be directly cooled by the exhaust water. We are exploiting the fact that there is a 625x multiplier for liquid:liquid heat transfer, but only a 1x multiplier for liquid:gas heat transfer. Even though the exhaust water only briefly exists as liquid, during that brief existence it is exchanging heat at 625x speed into the liquid lock, but the liquid lock itself only exchanges heat very slowly with the steam. The liquid lock should ideally be filled with Naphtha which has a very low thermal conductivity, but in practice Crude Oil or Petroleum work just fine too because heat transfer between Steam and Liquid is just that slow. But Naphtha does work better if you care to make some (but you can use Crude Oil for the bottom tile). You could also use Liquid Phosphorus if you feel like being different, it works well. Most Basic Build Now I'll move onto what is probably the most basic working build, a modified self-cooled Steam Turbine Volcano Tamer: This build can certainly benefit from improvements, but I show it as about the simplest possible implementation. As previously noted, the Autosweeper and Conveyor Loader can be made of Iron as they are cooled directly by the exhaust water. Meanwhile, we want to eliminate the need for a Steel Conveyor Shutoff - the 100 kg of Steel for a Shutoff is truly trifling but it's the principle of the matter, dammit! The solution here is to only trickle the refined Iron out. An Iron Volcano only produces Iron at a rate of about 0.5 kg/s during its active period so there is no good reason to release Iron from the build at a rate faster than that. A Timer Sensor set to 1/40 means that a 20 kg lump of iron will be released every 40 s, meaning Iron exits the system at a rate of 0.5 kg/s. This forces the Iron in the system to dwell for long enough to equalize temperature with the steam room. Generally, iron will exit this design at 125-160 C, which we can certainly improve upon! A note about Tempshift Plates: I would advise placing a single Tempshift Plate behind the volcano. It should be made of a terrible material, preferably Mafic Rock or Obsidian, this is adequate to instantly condense the molten Iron, but it ensures as much heat as possible is retained in the refined Iron debris and that limits the steam temperature surge during an eruption, using a high conductivity material will make the tamer work worse. It is also critical that the Tempshift Plate not transfer heat into the liquid lock or the tile above the liquid lock (where the exhaust water momentarily exists). So basically don't place Tempshift Plates on the left side of the Volcano. Also make sure no other building (like liquid bridge) can conduct heat from the steam into the liquid lock. A Bigger and Better Build This is a more realistic design which is definitely stable long term and lowers the Iron exit temperature to about 105 C. I move from 2 Steam Turbines to 3 Steam Turbines. It so happens that the most puny Iron Volcanoes actually can be tamed by 2 self-cooled Steam Turbines, but the more violent ones definitely require 3, so if you don't want to sit down and run the numbers, it's always safest to just use 3 Steam Turbines. I also add a secondary debris heat exchanger under the Iron Volcano. The exhaust water provides a great deal of surplus cooling above and beyond what the shipping machines require, and so I use a fairly typical "Door sandwich" to transfer heat from the secondary heat exchanger into the liquid lock, with a mostly precautionary temperature sensor which opens the door if the lock is too hot (which is unlikely). I also use a pair of Liquid Vents. One of them drips directly into the liquid lock, but the other drips onto the insulated tiles: that one is used for cooling the steam down if it exceeds 137 C. It will normally be disabled but I find this a useful measure to avoid "hot spots" forming in the steam chamber. The other overlays aren't doing anything fancy. You can also use an Iron Ore Aquatuner If you are a member of the cult of Aquatuners and firmly believe that a build without an Aquatuner is an unholy abomination, then an Aquatuner is easily integrated. As shown, you can simply make the Liquid Lock deeper and build an Iron Ore Aquatuner inside it where it will be cooled by the exhaust water allowing it to operate at a temperature well below 125 C. You need to take the automation signal which activates the Aquatuner and AND it with a Thermo Sensor set to say < 123 C as a failsafe against overheating the Liquid Lock. As long as there is a primary heat source to heat the steam to over 125 C and trigger the Steam Turbine(s) to activate you can run the Aquatuner, not nearly at full uptime but enough to do the basics like cooling the Steam Turbine and cooling the refined metal a bit extra. I personally would not recommend an Aquatuner though. Rejected "Oxygen Layer" Design Finally I want to share one of the inventions I rejected along the way, not because it didn't work, but because it wasn't as elegant as I would like: It's not very obvious, but the Autosweeper and the Conveyor Loader are actually sitting in a layer of oxygen gas trapped against the ceiling. If you've ever experienced a layer of low pressure oxygen gas blocking the Steam Turbine inlets in a not properly vacuumed out steam chamber you'll know exactly what I mean, but in this case a special place is provided for the oxygen layer to accumulate where it will not bother the Steam Turbines but provides very useful insulation for the Sweeper and Loader. In ONI gas:gas heat transfer is very slow (as it only has a 1x multiplier) but convection results in much higher practical heat transfer in gas, by convection, I mean tiles swapping places and carrying their heat with them, so even though the tiles aren't exchanging heat with each other very fast, they are rapidly swapping places so the heat is still moving around rapidly. But gases of different types are not allowed to swap places vertically except on the basis of density, so the Steam:Oxygen transition acts as an absolute barrier against convective heat transfer, heat in the steam only gets into the oxygen very slowly. The machines are cooled by the exhaust water from the Steam Turbines, which is Valved to 1000 g/s and runs through Radiant Pipes behind the oxygen layer. Furthermore, heat transfer between debris and oxygen gas is not very fast because it uses the lowest thermal conductivity and Oxygen is not very conductive, so storing 1000 C+ iron debris in oxygen is surprisingly manageable with the exhaust water cooling. This design actually does accomplish the purpose of allowing the use of common metal conveyor machines and it does so perfectly well. But I don't love it because it requires valving and the oxygen layer is arguably fickle, it actually does form very reliably but it can also be destroyed by construction mishaps. Basically compared with dripping the exhaust water into a liquid lock I find it too "busy". Naturally I also experimented with heavy gas insulation, like putting the Sweeper in a layer of chlorine gas in a put on the floor, which also can be made to work but suffers badly from disruption by stray liquids. However I feel that the gas layer approach, whether a light gas or heavy gas, has potential for using machines (including Steel or even Thermium) in an environment much too hot for them to exist, like an environment with rock gas or something, because it can easily allow for extremely high temperature deltas.
  8. There is clearly no such rule because smaller temperature changes are easily observed. If you aren't using Debug mode, then to easily see this you need to take advantage of the fact that the game displays temperatures between 0 and 1 C to 4 decimal places instead of 1 decimal place, then it can be seen that a 0 C tile can increase in temperature to 0.0001 C or whatever. With Debug mode enabled you can use the sample tool to see a high precision temperature for any tile at any temperature.
  9. Well in practice generally a temperature change smaller than 0.0001 K is not observed when the temperatures is in 3 digits. Which is 4 decimal places. So seems like a good starting point for where to limit things to avoid floating point wonkiness. For example, if the tool I'm using is correct (relative to the floating point system ONI uses), a 0.0001 K temperature change to something at around 300 K, can involve substantial inaccuracy, for example 273.15 is represented as 273.149993896484375 and 273.1501 is represented as 273.15008544921875, so what should be a 0.0001 K temperature change is actually 0.0000915 K, 9% error. Clearly this kind of temperature change is at the limit of 32 bit floating point precision and the real question is if the magic delta is a direct and natural consequence of the floating point calculations, or if there is some code which acts as a safeguard against going too deeply into the territory of floating point imprecision - but I'd expect it's the former case. I think you're thinking of 0.1 DTU? This can apply with Insulated Tiles, but it's trivially obvious that it's not the mechanism responsible for the magic delta. For example the heat transfer with an Insulated Tile should be identical whether it's made of Igneous Rock or Obsidian since both have identical thermal conductivity, but Igneous Rock has a magic delta 5x greater than Obsidian. So there is a temperature range where the Obsidian Insulated Tile is exchanging heat, but the Igneous Insulated Tile is not. This doesn't make sense in terms of the 0.1 DTU rule since both should exchange exactly the same amount of DTUs if they are the same temperature. In fact, we can determine that at the magic delta for a 20 C Igneous Rock Insulated Tile, an Obsidian Insulated Tile is transferring 8 DTU per tick and the Igneous Rock one should be too. I have seen it proposed that there is a minimum temperature change, perhaps the value of 0.0001 K, but testing readily shows this is is only the minimum temperature change at 3 digit temperatures. For example close to 0 K, a temperature change such as 0.000004 K can be observed (using Debug mode).
  10. I've done further testing in game and found something more complex. For example, when petroleum is on 0 C Igneous Insulated Tile, the magic temperature for zero heat transfer is 248.1 C. But on the other hand, if 0 C petroleum is on a hot Insulated Tile, the magic temperature (the hottest the insulated tile can be without heat transfer) is 496 C. On the third hand, if the Insulated Tile is at -272 C, the magic temperature is -268 C, a delta of only 4 C. The fact that the magic delta gets smaller as the insulated tile gets colder, suggests it's because the floating point system can more precisely represent values which are closer to 0. Hence the concept of the "magic delta" is broadly applicable for insulated tiles at normal temperatures, like normally they range from 15-45 C. But under highly abnormal conditions, like inside a liquid hydrogen chamber, the magic delta concept is no longer so applicable, an insulated tile at -250 C will gladly exchange heat with neighboring solid tiles. These are the numbers I measured in game with reference to Insulated Tile at 20 C in contact with a hotter liquid (altough the numbers for contact with a colder liquid are almost the same): Insulated Tile Material Magic Delta Magic Delta (Gas) Sedimentary 49.6 2.0 Obsidian 49.6 2.0 Mafic 99.3 4.0 Granite 115.4 4.6 Sandstone 136.8 5.5 Fossil 234.4 9.4 Igneous Rock 247.9 9.9 Ceramic 671.5 26.9 Isoresin 3792.0 151.7 Insulation N/A N/A As previously mentioned, this is only precise for Insulated Tile temperature of 20 C. It can be approximated that the magic delta doubles if the temperature (in Kelvin) doubles, and halves if the temperature is halved. For example if a Mafic Insulated Tile has a temperature of 313 C, it has a magic delta of 198 C, and if it has a temperature of -126 C, it has a magic delta of 49.6 C. An approximate formula for magic delta is: 1.692 * (T+273.15) * SHC / TC , where T, SHC and TC are the temperature, specific heat capacity and thermal conductivity of the Insulated Tile material. That formula can be used to extrapolate that the magic delta for gas with room temperature Insulation Insulated Tile with gas, is around 11 million C.
  11. This would actually be a useful thing to have on the wiki, if no-one beats me to it I'll derive the exact formula for the magic delta (which should be pretty simple) and make a chart for every relevant material.
  12. This is a common misconception because in the game the thermal conductivity of abyssalite is rounded to 0, but it's actually 0.00001. In fact abyssalite usually conducts more heat than insulated tiles, usually about 2-30x as much depending on the insulation material and exact elements involved. (in particular, Insulated Tiles are much betterer at dealing with high conductivity gases, they are still better when dealing with low conductivity gases but the disparity is smaller) The game does not perform heat transfers which it considers too small to be significant, so for example there is usually truly zero heat transfer between two abyssalite tiles, or two insulated tiles. There also may be zero heat transfer between an insulated tile and solid or liquid (putting aside flaking) but it is very peculiar to the exact materials and temperature delta involved. Gas will nearly always exchange heat with Insulated Tiles (unless they are made of Insulation) if there is a significant temperature delta, and hydrogen and steam are particularly prone to causing issues as they have very high thermal conductivity for gases.
  13. This game has endless surprises. I have known for a very long time that heat rises (and spreads horizontally very fast) in gases, but honestly wasn't sure how exactly it worked in liquids, except knowing that the 1 degree rule applies strongly to liquids except when heat is rising. The surprise is that the devs implemented "heat rising" in liquid in a completely different way to gases. Left hand side: 100 kg Oxygen Gas tiles at 0 or 100 C. Right hand side: 1100 kg Ethanol tiles at 0 or 70 C. Top: Hot above cold (heat rising). Bottom: Cold above hot (heat sinking). Convection in gases works in the following way. Gas tiles swap places randomly and rapidly horizontally (there used to be horizontally temperature smearing, but that was removed to fix a bug), and they swap places vertically occasionally but only if a hotter tile is beneath a colder tile. This is a form of true convection as the heat is being carried by the movement of the fluid. (the game doesn't form convection currents though and it's not based on density, hence it can't really be described as natural/free convection, but it is a form of heat convection, just with the motion being driven by something like macroscopic "Brownian motion" with a bias for heat rising) Heat rising in liquids works in the following way: Normal conduction always applies horizontally, there is no random tile swapping. When a hotter tile is under a cooler tile, every few ticks they instantaneously and fully equalize in temperature (in a heat-conserving way), which bypasses thermal conductivity and the 1 degree rule, as far as I can tell there is no "convection" mechanism, as in the tile is not actually moving, only the heat is moving. Though this is clearly meant to simulate convection in liquids. Because liquids have quite high thermal conductivity in all directions, the most practically significant effect of this comes from the 1 degree rule. In both gases and liquids the 1 degree rule is ignored when the "hot plate" is at the bottom of the room/pool, this causes the entire vertical column above the hot plate to reach the same temperature. In gas, the horizontal convection will also cause the tiles to reach the same temperature horizontally, but in liquid the 1 degree rule means that the temperature will not equalize horizontally. Another thing to note is that vertical heat conduction in liquids is potentially astonishingly fast, in the video did you see how quickly the heat rose up the column? Though the potential heat transfer rate is so high I'm not sure what could actually make use of it.
  14. Doors actually are two tiles, but the door object only shows the temperature for the left or bottom tile (and when the door is opened, the tiles are deleted and the temperature is averaged, and when it's closed the tiles are created at the same temperature, but can then diverge). It is straightforward to demonstrate that doors have two tiles which can have different temperatures. Make a door, and build a one tile object inside each tile of the door (i.e. an automation wire, or whatever, anything really), put heat across the door so there should be a gradient (i.e. one side hot, the other side cold), and check the temperature of each object. It can be observed that they have different temperatures, hence the different tiles of a door do have different temperatures. And an even simpler way, is to just click twice on each side of the door, so you see the underlying tile, and it will have its own temperature under the properties tab. I am not doubting that in some cases doors make better heat exchangers, but if so it's probably 100% due to them having different material properties, i.e. a Mechanical Airlock consists of 2 200 kg Tiles, wheras a Metal Tile is a 100 kg Tile, and higher mass (heat capacity) can make for more effective heat exchange. You can also verify this hypothesis by using Sandbox Mode brush tool to compare a Steel Mechanical Airlock with a pair of 200 kg natural tiles made of Steel, they should be precisely identical in performance. Or you could compare the performance of a Manual Airlock, with a pair of Hydroponic Tiles of the same material. edit: These rules don't apply to Pneumatic Doors or open Airlocks. I don't know what rules apply to Pneumatic doors, but they act neither like tiles nor buildings, having somewhat slower heat exchange than comparable buildings and much slower than tiles.
  15. Best way to kill dupes... Make a small room with hamster wheels and a Pneumatic Door. On the door, there are door restrictions, set the door restrictions to allow the doomed dupes to come in, but not leave. Then order them in (tell them to move into the room). Dupes obey door restrictions even if it is life or death, and they will run on the hamster wheels until they starve to death. Of course, the hamster wheels are entirely optional, you could just use 2 tile room behind an airlock where they will quickly suffocate, but may as well get some use of the last cycles of their life.