Acoustic loads — SPL, octave band analysis: Visual walkthrough

3.6.12 · D2 · Physics › Spacecraft Structures & Systems Engineering › Acoustic loads — SPL, octave band analysis

Yeh page poori acoustic-loads ki kahani ko sirf ek hilte hue air molecule se dobara banata hai. Har symbol ko use karne se pehle hum usse earn karenge: pressure, RMS, decibel, reference number $20 μ Pa$ , aur aakhir mein woh rule jo kai octave bands ko ek Overall SPL mein silta hai. Pictures follow karo — woh derivation carry karte hain.

Parent: Acoustic Loads — SPL, Octave Band Analysis.

Step 1 — Sound ka "pressure" actually hota kya hai

WHAT. Air normally har surface par ek steady atmospheric pressure se push karta hai, roughly $p_{0} = 1 0^{5} Pa$ (pascals — newtons of push per square metre). Ek sound wave ek tiny extra push-and-pull hai jo us steady value ke upar sawaar hoti hai. Is wobble ko hum $p (t)$ kehte hain: yeh thodi positive (compression) hoti hai, phir thodi negative (rarefaction), baar baar.

WHY. Isse pehle ki hum "kitna loud hai" measure kar sakein, hume woh cheez name karni hogi jo hum measure karte hain. Loudness total pressure $p_{0} + p (t)$ nahi hai — ek microphone constant $p_{0}$ ko ignore karta hai aur sirf wobble $p (t)$ report karta hai. Toh $p (t)$ humara raw ingredient hai.

PICTURE. Figure mein flat line $p_{0}$ hai; wavy line true pressure hai. Unke beech ka shaded gap — upar aur neeche swing karta hua — woh $p (t)$ hai, yaani sound.

Figure — Acoustic loads — SPL, octave band analysis

Step 2 — Hum $p (t)$ ko seedha average kyun nahi kar sakte: RMS ki entry

WHAT. Hum chahte hain ek number jo bataye "wobble kitna bada hai." Obvious idea — $p (t)$ ko time ke saath average karna — fail ho jaata hai: wave utna hi waqt positive aur negative mein spend karti hai, isliye uska average exactly zero hota hai. Iske bajaye hum pehle square karte hain (har value positive ho jaati hai), us ka average lete hain, phir square root lete hain. Yeh hai root-mean-square, likha jaata hai $p_{rms}$ .

WHY square? Do wajahaat jo actually ek hi hain. (1) Squaring sign khatam kar deta hai, toh cancellation wave ko chupa nahi sakti. (2) Ek wave jo energy carry karta hai woh pressure squared ke proportional hai — isliye $p^{2}$ ka mean physically meaningful average hai, $p$ ka mean nahi.

PICTURE. Figure mein raw wave dikhti hai (zero cross karti hui), phir squared wave $p^{2}$ (sab zero ke upar, ek hump train), phir $p^{2}$ ke mean par flat dashed line; $p_{rms}$ us height ka square root hai.

Step 3 — Pressure se energy flow tak (intensity)

WHAT. Ek wave sirf air ko jagah par squeeze nahi karta; woh molecules ko along bhi nudge karta hai, unhe ek velocity $v (t)$ deta hai. Ek surface se guzarti power per square metre intensity $I$ hai. Ek plane wave ke liye, pressure aur velocity ek number se lock hote hain — acoustic impedance $ρ c$ .

WHY this tool? Hume intensity chahiye kyunki energy woh hai jo add hoti hai jab kai sources ya kai frequency bands combine hote hain (Steps 6–8 mein aata hai). Pressure akela cleanly add nahi hota; intensity hoti hai. Link $p = ρ c v$ hi hai jo hume pressure ko energy ke saath trade karne deta hai — $ρ c$ medium ka natural "stiffness × speed" kyun hai yeh dekho 2.5.8 Acoustic Impedance and Transmission mein.

PICTURE. Figure mein air ka ek slab hai: ek pressure arrow right push kar raha hai, ek velocity arrow right move kar raha hai, aur product $p \cdot v$ face se stream ho rahi energy ke roop mein shaded hai. $p$ -vs- $v$ line ki slope exactly $ρ c$ hai.

Step 4 — Loudness ko logarithm ki zaroorat kyun hai

WHAT. Sabse quiet audible sound ka $p_{rms} \approx 20 μ Pa$ hai (woh hai $20 \times 1 0^{- 6}$ Pa). Full blast par ek jet roughly $200$ Pa ke aas paas hai. Yeh ek-karod-to-one ka span hai. "0.00002 Pa … 200 Pa" likhna bekaar hai. Toh hum ise $lo g_{10}$ se compress karte hain: logarithm poochta hai "ten ki kitni power yeh ratio deti hai?" aur multiplication ko addition mein badal deta hai.

WHY log aur square root nahi, maano? Square root range ko thoda hi shrink karta hai ( $1 0^{7} \approx 3162$ , phir bhi huge). Logarithm kisi bhi number of zeros ko ek chhote count mein shrink karta hai: $lo g_{10} (1 0^{7}) = 7$ . Kyunki hamare kaan khud roughly ratios par respond karte hain (pressure double karna ek fixed step jaisa lagta hai chahe soft se shuru karo ya loud se), log perception se bhi match karta hai.

PICTURE. Left panel: ek linear axis jahan whisper invisibly zero ke paas hai aur jet page se bahar run kar raha hai. Right panel: same points log axis par, ab $- 6$ se $+ 2$ exponent mein evenly spread — readable.

Step 5 — SPL formula assemble karna

WHAT. Ab hum Sound Pressure Level (SPL) define karte hain. Hum apni sound ki intensity ko hearing-threshold intensity $I_{ref} = 1 0^{- 12} W/m^{2}$ se compare karte hain, log lete hain, aur decibels banane ke liye 10 se scale karte hain. Kyunki $I \propto p^{2}$ hai, pressure form mein 10 ki jagah 20 aata hai.

WHY 10, aur kyun 20 ban jaata hai? Decibel mein "deci" ka matlab hai hum bel ke tenths use karte hain, isliye intensity level $10 lo g_{10} (I / I_{ref})$ hai. Intensity ki jagah pressure use karne par $I \propto p^{2}$ use hota hai, aur $lo g_{10} (x^{2}) = 2 lo g_{10} (x)$ — woh factor of 2, 10 ke saath ride karta hai aur 20 deta hai. Reference $p_{ref} = 20 μ Pa$ arbitrary nahi hai: yeh exactly woh pressure hai jo $I_{ref}$ ko Step 3 ke through match karta hai, $p_{ref} = I_{ref} ρ c$ .

PICTURE. Ek flow arrow: raw $p_{rms}$ → $p_{ref}$ se divide karo → $lo g_{10}$ lo → 20 se multiply karo → dB. Har box annotated hai ki usse kya nikalta hai. Ek side ruler landmark levels dikhata hai: 0 dB threshold, 60 dB baat karna, 120 dB pain, 150 dB fairing.

Step 6 — Decibels ordinary numbers ki tarah kyun kabhi add nahi hote

WHAT. Do engines mein se har ek 145 dB par hain toh woh 290 dB nahi banate. Decibels energy ke logs hain, aur energies add karne ke liye pehle log se bahar aana padhta hai, real intensities add karo, phir wapas andar jaao.

WHY intensities add hoti hain pressures nahi uncorrelated sources ke liye? Do uncorrelated sources mein random relative phase hoti hai. Jab tum summed pressure square karte ho, $⟨(p_{1} + p_{2})^{2} ⟩ = ⟨ p_{1}^{2} ⟩ + ⟨ p_{2}^{2} ⟩ + 2 ⟨ p_{1} p_{2} ⟩$ , aur woh aakhri cross-term zero par average ho jaata hai kyunki phases independently wander karte hain. Jo bachta hai woh $⟨ p_{1}^{2} ⟩ + ⟨ p_{2}^{2} ⟩$ hai — do intensities, added. Toh energy woh currency hai jo sum hoti hai.

PICTURE. Left: do random waves aur unka sum; cross-term area half positive, half negative draw kiya gaya hai, cancel hota hua. Right: do intensity bars ek talle bar mein stack hote hue, dB tick dikhata hai ki stack sirf $10 lo g_{10} 2 \approx 3$ dB barhta hai.

Step 7 — Spectrum ko octave bands mein slice karna

WHAT. Real launch noise ek tone nahi hota; yeh energy frequencies ke across spread hoti hai. Hum us continuous spectrum ko octave bands mein kaatte hain — har band ek 2:1 frequency ratio, standard values $31.5, 63, 125, 250, 500, 1000, 2000, 4000, 8000$ Hz par centred.

WHY $2$ edges? Hum chahte hain ki band logarithmically centred ho, kyunki frequency perception (aur structural resonances) ratio-based hain. $lo g f_{c} = \frac{1}{2} (lo g f_{low} + lo g f_{high})$ demand karne par $f_{c}^{2} = f_{low} f_{high}$ force hota hai. Octave rule $f_{high} = 2 f_{low}$ ke saath combine karo toh milta hai $f_{low} = f_{c} / 2$ , $f_{high} = 2 f_{c}$ . Yahan se $2$ aata hai — koi magic nahi, bas " $f_{c}$ ko geometric middle mein rakho."

PICTURE. Ek log-frequency axis standard centres par tick marks ke saath; har centre ke upar ek bracket $f_{c} / 2$ se $2 f_{c}$ tak run karta hai, aur har bracket log axis par same width ka hai (visually "octave" ka yehi matlab hai).

Step 8 — Bands ko ek number mein wapas silna (OASPL)

WHAT. Har band ka level $L_{i}$ (dB) given ho toh, Overall SPL na sabse loud band hai aur na dB values ka sum. Kyunki bands non-overlapping frequencies cover karte hain, unki energies simply add hoti hain — exactly Step 6 ki logic, ab $N$ pieces ke liye.

WHY $1 0^{L_{i} /10}$ sum karo? Har band ki intensity $I_{i} = I_{ref} 1 0^{L_{i} /10}$ hai (log undo karo). Non-overlapping bands interfere nahi kar sakte, isliye total energy $\sum I_{i}$ hai. Ise $10 lo g_{10}$ se wapas daalo aur $I_{ref}$ cancel ho jaata hai, ek clean sum-of-tens formula bacha ke.

PICTURE. Paanch vertical bars, ek har band ke liye, heights = $1 0^{L_{i} /10}$ (relative intensity). Woh ek tall bar mein stack hote hain; right side par ek dB scale dikhata hai ki stack 143.4 dB par land karta hai — sabse tall single bar (140 dB) se barely upar, kyunki 500 Hz band dominate karta hai.

Step 9 — Edge aur degenerate cases (reader ko kabhi stranded mat chhodho)

WHAT & WHY & PICTURE (teen quick cases):

Silence, $p_{rms} \to 0$ . Tab $p_{rms} / p_{ref} \to 0$ aur $lo g_{10} (0) \to - \infty$ : SPL $- \infty$ dB tak dive karta hai. Physically: perfect silence threshold se "infinitely below" hai. Picture mein SPL curve bottom se plunge karta hua dikhta hai jab pressure zero approach karta hai.
Exactly at threshold, $p_{rms} = p_{ref}$ . Ratio $= 1$ , $lo g_{10} 1 = 0$ , isliye SPL $= 0$ dB. Zero dB silence nahi hai — yeh sabse faint audible tone hai. Curve yahan axis cross karta hai.
OASPL mein ek band dominate kare. Agar band $k$ baaki sab se $\geq 10$ dB upar hai, $1 0^{L_{k} /10}$ sum ko swamp karta hai aur $L_{OASPL} \approx L_{k}$ hota hai. Picture: ek giant bar aur negligible others — overall level peak par "clamp" ho jaata hai. Isliye sabse loud band dhundhna (aur kya woh ek resonance hit karta hai) hi poora game hai.

Ek-picture summary

Ek diagram, poori chain: ek hilta molecule → sound pressure $p (t)$ → RMS → intensity $\propto p^{2}$ → log-compress → SPL in dB → octave bands mein slice karo → energy-sum wapas OASPL tak.

Recall Feynman retelling — ise ek story ki tarah bolo

Air normally sab kuch par ek steady weight se push kar rahi hoti hai. Ek sound us push par ek tiny extra jiggle hota hai. Agar main jiggle ko average karne ki koshish karun toh zero milta hai, kyunki woh utna hi upar jaata hai jitna neeche — toh main pehle ise square karta hun (jo actually energy bhi hoti hai), average karta hun, aur wapas square-root leta hun; yahi RMS jiggle hai. Ek pure tone ke liye squared wave apne peak ke aadhe ke around hover karta hai (kyunki $sin^{2}$ half tak average hota hai), toh RMS peak ko $2$ se divide karke nikalta hai. Woh jiggle energy carry karta hai, aur ek wall se kitni energy stream hoti hai woh air ki khud ki stiffness-times-speed par depend karta hai, ek number jise $ρ c$ kehte hain; energy jiggle-squared ki tarah jaati hai. Lekin quiet aur loud sounds mein das lakh ka antar hota hai, toh main inhe ek ruler par nahi likh sakta — main logarithm leta hun, jo sirf zeros count karta hai, aur main sabse faint sound se compare karta hun jo ek insaan sun sakta hai. Bees se scale karo (das decibel ke liye, do kyunki energy pressure-squared hai) aur main decibel pa leta hun, SPL. Decibels logs hain, toh main inhe kabhi seedha add nahi kar sakta: sounds combine karne ke liye main log se real energy par aata hun, energies add karta hun, wapas jaata hun. Launch noise kai pitches par spread hoti hai, isliye main spectrum ko octaves mein cut karta hun — har ek frequency ka doubling, log ruler par middle mein centred, isliye edges par woh $2$ aata hai. Aakhir mein, ek grand total ke liye main har band ki energy add karta hun aur ek baar log leta hun: yahi OASPL hai. Aur punchline — kyunki yeh energy-adding hai, sabse loud band almost hamesha jeet jaata hai; overall level choir ki sabse badi awaaz se barely upar uthta hai.

Recall Quick self-check

SPL mein 10 nahi 20 kyun? ::: Intensity level 10 use karta hai; pressure squared enter karta hai, aur $lo g (p^{2}) = 2 lo g p$ , toh 10×2 = 20. Sine $p_{rms} = \overset{p}{^} / 2$ kyun deta hai? ::: Kyunki $sin^{2}$ time-average ho kar $\frac{1}{2}$ banta hai ( $sin^{2} = \frac{1}{2} (1 - cos 2 t)$ use karte hue, jiska $cos$ part zero average karta hai), toh $p_{rms} = \frac{1}{2} \overset{p}{^}$ . Uncorrelated sources ke liye intensities add kyun hoti hain, pressures nahi? ::: Pressure cross-term $2 ⟨ p_{1} p_{2} ⟩$ random phase par zero average karta hai, $⟨ p_{1}^{2} ⟩ + ⟨ p_{2}^{2} ⟩$ — yaani intensities — bacha ke. Jab ek band baaki se 10 dB upar ho toh OASPL kya hoga? ::: Approximately us band ka apna hi level; woh energy sum dominate karta hai. $p_{rms} = p_{ref}$ ke corresponding SPL kya hai? ::: Exactly 0 dB (ratio 1 hai, $lo g_{10} 1 = 0$ ).

Related environments to keep in view: 3.6.13 Shock Loads and SRS, 3.6.14 Combined Environmental Testing.