I have a setup with 4 Combustion engines cooled by liquiduct attached to an Aqueous Accumulator. The Accumulator is connected to liquiduct at 4 faces, the max possible as it needs 2 faces to be adjacent to water. When the engines reach ~510 C, they start to lose water slowly. From what I've read, 1 AA should service 4 CE's through liquiduct no problem. What's the issue? Can screenshot if necessary
Combustion Engine Cooling problem with AA
- Thread starter Mahoka
- Start date
-
The FTB Forum is now read-only, and is here as an archive. To participate in our community discussions, please join our Discord! https://ftb.team/discord
AA can and does service 4 combustion engines with no problem.
Maybe a screenshot would be a good idea, because I'm not sure I understand when you say "The accumulator is connected to liquiduct at 4 faces"... does that mean what I think? You don't need to do that. Just connect a liquiduct to one face, and then further down, branch off into each engine.
Maybe a screenshot would be a good idea, because I'm not sure I understand when you say "The accumulator is connected to liquiduct at 4 faces"... does that mean what I think? You don't need to do that. Just connect a liquiduct to one face, and then further down, branch off into each engine.
Liquiducts can pass an infinite amount of liquid, but each connection is limited to a certain mb/s input/output. I added more connections to be sure this was not the limiting factor, as I already checked my accumulator and it had enough water inside it. I've reduced it to one face now like you said, and I'm going to wait till they heat up and observe what happens. I'll post a screenshot if it fails again.
Well, it's working now. The only difference between my first setup (before the multiple faces in liquiducts) and this one is that the AA used to be adjacent to one of the engines, and now it is only touching the liquiduct, well, thank you for the help
Aqueous Accumulators can produce half a bucket of water every second, roughly equal to 25 mB/t. Combustion Engines at full operating temperature consume 1 mB/MJ, and thus their water consumption rate depends on fuel. At maximum, they will consume 6 mB of water per tick per engine.
The tricky part is the liquiducts. Liquiduct connections can move 100 mB/t of water maximum (other fluids have different maximum fluid transfer rates depending on viscosity) -- but their current throughput is dependent on capacity of the liquiduct network. If the liquiduct is less than three quarters full, it will not get full throughput, and if it's less than a quarter full, each output connection will only get half of its potential throughput. This is especially tricky for long liquiduct runs, as internal capacity is determined by the number of liquiduct blocks in the network.
EDIT: [strike]in passive mode, and 200 mB/t in forced extraction mode[/strike] Forced extraction mode numbers are either outdated or based on a misunderstanding; I can't find code or any practical implementation that supports it. Liquiduct throughput looks to be the same regardless of mode.
The tricky part is the liquiducts. Liquiduct connections can move 100 mB/t of water maximum (other fluids have different maximum fluid transfer rates depending on viscosity) -- but their current throughput is dependent on capacity of the liquiduct network. If the liquiduct is less than three quarters full, it will not get full throughput, and if it's less than a quarter full, each output connection will only get half of its potential throughput. This is especially tricky for long liquiduct runs, as internal capacity is determined by the number of liquiduct blocks in the network.
EDIT: [strike]in passive mode, and 200 mB/t in forced extraction mode[/strike] Forced extraction mode numbers are either outdated or based on a misunderstanding; I can't find code or any practical implementation that supports it. Liquiduct throughput looks to be the same regardless of mode.
What do you mean by "less than x full"? Is this determined by the output/input pressure? i.e. If you have 200 mB/t from forced extraction mode and 34 combustion engines running off fuel at max temp (34x6 = 204 mB/t), what will happen? Will the engines eventually run out of water?
Is this problem easily fixed by adding another forced extraction connection to the network, parallel to the first?
Is this problem easily fixed by adding another forced extraction connection to the network, parallel to the first?
I stand corrected after some testing: throughput for water is the same regardless of the liquiduct mode, at 100 mB/t when transporting water.
And then boom.
In situations where the liquid consumption rate is higher than 50 mB/t, the precise nature of the system can be more relevant.
Note that, while this emulates total pressure for the system, it's a fairly simplified version of such things. There are no tests for bottlenecks between source inputs and destination outputs, for example.
With the above caveat, it ends up being a 17 engines to hit 102 MJ/t and thus 102 mB/t. The typical behavior you'll see is all of the engines running up smoothly to hit 510 degrees Celcius, after which they begin to drain water from the liquiduct network. The engine coolant tanks will stay full for a short time period, depending on the size of the liquiduct network. Because each engine only consume 6 mB/t, even the minimum output rate of 50 mB/t per output connection is still going to outpace the burn rate. This reduction in throughput isn't exacty pressure. Once the liquiduct network is extremely low -- or if it were empty to start with -- then individual combustion engine tanks will slowly empty, and once they empty the combustion engine heat will go up.
And then boom.
In situations where the liquid consumption rate is higher than 50 mB/t, the precise nature of the system can be more relevant.
Note that, while this emulates total pressure for the system, it's a fairly simplified version of such things. There are no tests for bottlenecks between source inputs and destination outputs, for example.
With my apologies for the confusion about extraction- and normal-mode connections, yes. Adding another input, as long as that input is attached to the same network, will add directly to the liquiduct's internal tank. That connection can sit directly in-line with the existing one, though, and you won't need to run another liquiduct line : again, liquiducts only check for throughput at inputs and outputs.
Another random thing to check is to make sure the redstone control of the AA is disabled. I almost had a boiler explode once. I had the power switch to activate my steam engines run by my AA. When I turned the engines on, it disabled the AA 
I luckily caught it in time.
I luckily caught it in time.
Another random thing to check is to make sure the redstone control of the AA is disabled. I almost had a boiler explode once. I had the power switch to activate my steam engines run by my AA. When I turned the engines on, it disabled the AA
I luckily caught it in time.
Yeah well maybe if you weren't AlwaysGoofingOff...
(see what I did thur?)Hmm. Well your post is kinda confusing, but I'll take ur word for it.
It would seem to me that its based on pressure, but pressure is consistent throughout the liquiduct, i.e. each machine will receive the same pressure, regardless of its position in the network. There is no internal bottle necking, which makes organizing liquids a lot easier than waterproof piping.
It would seem to me that its based on pressure, but pressure is consistent throughout the liquiduct, i.e. each machine will receive the same pressure, regardless of its position in the network. There is no internal bottle necking, which makes organizing liquids a lot easier than waterproof piping.
Sorry. For most people and for any imaginable circumstance involving combustion engines, all that matters is that the number of inputs times the throughput per input (with water, 25 mB/t from an Aqueous Accumulator, or 100 mB/t from a full tank) compared to the consumption rate.
For very high-consumption machines, or some very unusual situations where a tank is used for filtering, there are some very rare situations where you'd want to make sure the liquiducts themselves were full before turning on the machine, because the output would be or remain slow. These are not likely to occur, or matter much where they occur, though.
For very high-consumption machines, or some very unusual situations where a tank is used for filtering, there are some very rare situations where you'd want to make sure the liquiducts themselves were full before turning on the machine, because the output would be or remain slow. These are not likely to occur, or matter much where they occur, though.