Wikipedia:Reference desk/Archives/Science/2025 March 22

Science desk
< March 21<< Feb | March | Apr >>March 23 >
Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.


March 22

Isp calculations

I've been working on a tool to calculate delta-v, and I got a little stuck. Specifically, how do you calculate specific impulse of a rocket when you have multiple engines with different specific impulse and thrust firing at the same time (example: 4x RS-25 + 2 SRBs)? Assume constant thrust, no atmosphere, no gravity. Stoplookin9 Hey there! Send me a message! 02:13, 22 March 2025 (UTC)[reply]

The acceleration of the rocket equals its mass divided by the thrust. The combined thrust is constant (assuming no engine runs out of fuel); it equals the sum of the thrusts of all engines. Let denote the thrust of engine E. The total thrust is then given by the sum of all individual thrusts: The difficulty of the problem is that the acceleration is not constant. The mass decreases as rocket fuel is consumed, which means that the acceleration increases. To solve this requires to use some integral calculus. We need to know the total mass of the system at the start; let's call it . How does it decrease as fuel is burned? Let denote the the specific impulse of engine E. The contribution of this engine to the rate of fuel mass consumption equals Sum this quantity over all engines to get the total mass flow rate . After a time has elapsed, the mass of the whole system will have decreased to So the acceleration at time after the start equals To get after a burn time we need to integrate this:
The function in this formula is the natural logarithm, so, for example,  ​‑‑Lambiam 11:49, 22 March 2025 (UTC)[reply]
Specific pulse is the ration of the trust to the mass flow of the propellant. You can do the simple calculation knowing thrusts and specific pulses of all engines. Ruslik_Zero 20:02, 22 March 2025 (UTC)[reply]
This only works if the mass of the fuel consumed during the burn (in the formula above) is insignificant, compared to the total mass. Writing we can expand as a Taylor series:
 ​‑‑Lambiam 07:48, 23 March 2025 (UTC)[reply]
It will work in any case:
.
All these engine parameters (thrusts and specific fuel consumptions) are just constants. Ruslik_Zero 19:05, 23 March 2025 (UTC)[reply]
Actually in your formula . Ruslik_Zero 19:10, 23 March 2025 (UTC)[reply]
How do you get from to ?  ​‑‑Lambiam 09:19, 24 March 2025 (UTC)[reply]
By tsiolkovsky rocket equation as usual. Ruslik_Zero 19:52, 24 March 2025 (UTC)[reply]
Tsiolkovsky's formula is the answer I gave; is the remaining mass after a burn time Tsiolkovsky's formula assumes that all fuel is consumed. In the case of multiple engines with different characteristics, I don't think we can assume all burn out at the same time, and then the formula no longer works. You need to calculate to the first burn-out, then add the from that point till the next burn-out, and so on.  ​‑‑Lambiam 23:23, 24 March 2025 (UTC)[reply]

What is the correct name for prompts used for AI generation

Is it AI-generation prompt or AI-generated prompt Trade (talk) 11:43, 22 March 2025 (UTC)[reply]

Neither of the two sounds natural to me. If it is clear that the context is generative AI, just "prompt" will generally do just fine. If, in the context, you need some attribute to distinguish it from other kinds of prompts, you can use "GenAI prompt"[1] (or possibly, depending on the audience, the longer term "generative-AI prompt").  ​‑‑Lambiam 12:00, 22 March 2025 (UTC)[reply]
"AI-generated prompt" is the least correct of the options the OP and Lambiam have proposed. It means that the prompt is generated by AI, not that the prompt is used to have AI generate something. DMacks (talk) 12:18, 22 March 2025 (UTC)[reply]
"AI prompt" is used in documentation for ChatGPT, DeepSeek, and DALL-E. The word "generative" has special meaning in AI. You can't just toss it around willy-nilly. 12.116.29.106 (talk) 17:34, 25 March 2025 (UTC)[reply]
Uses material from the Wikipedia article Wikipedia:Reference desk/Archives/Science/2025 March 22, released under the CC BY-SA 4.0 license.