Visual walkthrough — Acoustic modes in combustion chamber — cause of instability

3.3.33 · D2 · Physics › Rocket Propulsion › Acoustic modes in combustion chamber — cause of instability

Hum ek hi sawaal ka jawab pictures mein de rahe hain:

Hot gas ek tube mein special frequencies kyun choose karta hai, aur burning fuel unme se ek ko itna loud kyun bana sakta hai ki sab kuch blast ho jaye?

---

Step 0 — Teen words jo pehle samajhne zaroori hain

Kisi bhi symbol se pehle, teen plain-word ideas, har ek ke saath aage aane waali picture reference.

Neeche sab kuch sirf inheen teen cheezon se bana hai.

---

Step 1 — Ek wobble jo apne neighbour ko dhakelta hai: wave equation

KYA. Hum gas ke ek patle slice ko dekhte hain. Jab uske left par pressure uske right se zyada hota hai, slice rightward push hoti hai. Woh push change karta hai ki agli slice kitni squish hai, jo uska pressure change karta hai, aur yahi silsila chalta rehta hai.

YEH TOOL KYUN. Hume ek rule chahiye jo kahe "gas ki acceleration pressure ke difference se hoti hai." Woh difference-of-a-difference bilkul wahi hai jo symbol ${\nabla }^2$ (aur 1-D mein, second derivative ${\partial }^2/\partial x^2$) measure karta hai — pressure shape ki curvature. Curvature sahi tool hai kyunki pressure ki straight ramp uniform push deti hai (koi net squeezing change nahi), sirf pressure profile mein bend gas ko pile up ya spread out karata hai.

"Gas accelerates because of pressure curvature" ko symbols mein likhne par wave equation milta hai:

$\underset{\text{ek point par squeeze time mein kitni tezi se speed up hoti hai}}{\underbrace{\frac{{\partial }^2{p}^{\prime }}{\partial t^2}}}=\underset{\begin{array}{c}\text{hand-off ki speed,}\\ \text{squared}\end{array}}{\underbrace{c^2}}\underset{\text{tube ke saath squeeze ki curvature}}{\underbrace{\frac{{\partial }^2{p}^{\prime }}{\partial x^2}}}$

• $\frac{{\partial }^2{p}^{\prime }}{\partial t^2}$ — symbol $\partial /\partial t$ matlab hai "time mein change ki rate, spot fixed rakhke"; do baar karne par us spot par pressure ki acceleration milti hai.
• $\frac{{\partial }^2{p}^{\prime }}{\partial x^2}$ — $\partial /\partial x$ matlab hai "tube ke saath step karte waqt change ki rate"; do baar karne par curvature milti hai — pressure graph kitna bend karta hai.
• $c^2$ — conversion factor. Bada $c$ (hotter gas) matlab wahi bend faster response produce karta hai.

---

Step 2 — Resonance ek hi pitch par hum karti hai: space aur time alag karna

KYA. Hum guess karte hain ki ek resonance ek fixed shape rakhti hai aur bas ek steady rhythm mein saans andar-bahar leti hai. Toh hum $p^{\prime }$ ko "sirf position par depend karne wali shape" times "sirf time par depend karne wali wobble" mein split karte hain.

KYUN. Apni lowest note par vibrate karta guitar string apni shape nahi badalta — middle bas upar-neeche jaata hai. Yahan bhi waisa hi hai: pattern frozen hai, sirf iska amplitude pulse karta hai. Ise product ke roop mein likhne se hum shape aur rhythm ko alag-alag solve kar sakte hain.

$p^{\prime }(x,t)=\underset{\begin{array}{c}\text{tube ke saath}\\ \text{frozen shape}\end{array}}{\underbrace{P(x)}}\times \underset{\begin{array}{c}\text{time mein andar-bahar}\\ \text{saans lena}\end{array}}{\underbrace{\cos (\omega t)}}$

• $P(x)$ — standing pattern: batata hai ki har position $x$ par wobble kitni badi hai. Yeh kabhi move nahi karta.
• $\cos (\omega t)$ — ek number jo hamesha ke liye $+1$ aur $-1$ ke beech smoothly swing karta hai. $\cos$ kisi bhi evenly repeating cheez ke liye tool hai, kyunki yeh steady speed par circle mein ghoom rahe point ki shadow hi hai.
• $\omega$ (Greek "omega") — angular frequency: woh circle kitni tezi se spin karta hai, radians per second mein. Bada $\omega$ = fast breathing = high pitch.

Yeh guess wave equation mein plug karo. $\cos (\omega t)$ ka time-derivative do baar lene par factor $-{\omega }^2$ neeche aa jaata hai; space part sirf $P(x)$ ko touch karta hai:

$-{\omega }^{2}P=c^2{P}^{\prime \prime }⟹\underset{\text{shape equation}}{\underbrace{P^{\prime \prime }+k^{2}P=0}},\underset{\text{wavenumber}}{\underbrace{k=\frac{\omega }{c}}}$

• $P^{\prime \prime }$ — frozen shape ki curvature.
• $k$ — wavenumber: ek metre mein kitne radians of wave fit hote hain. Kyunki $k=\omega /c$, fast wobble ($\omega$) slow gas mein (small $c$) zyada waves thoos deta hai.

---

Step 3 — Walls decide karti hain ki kaunsa cosine survive karta hai

KYA. Chamber ke ends hain. Har end par hume ek physical rule satisfy karni hai, aur sirf special cosines ise mante hain.

"Velocity = 0 at a rigid wall" rule kyun. Gas solid wall se through nahi flow kar sakta. No flow matlab gas velocity wall ke bilkul paas zero hai. Ab yeh key mechanical fact hai: gas velocity pressure ke slope se drive hoti hai (pressure difference gas ko push karta hai). Zero velocity matlab pressure ka zero slope — wave ki flat top ya flat bottom. Wave ka flat-topped point uska peak hai, yaani pressure antinode, node NAHI.

Hume aisi shape chahiye jiske dono ends par zero slope ho $x=0$ aur $x=L$ par:

$P(x)=A\cos \left(\underset{\text{kitna tight}}{\underbrace{k}}x\right),\text{slope }{P}^{\prime }(x)=-Ak\sin (kx)$

• $x=0$ par: $\sin (0)=0$ ✓ — cosine automatically flat shuru hota hai.
• $x=L$ par: hume $\sin (kL)=0$ chahiye. Yeh tabhi hota hai jab $kL$ half-turns ki poori number ho:

$kL=n\pi ⟹k_n=\frac{n\pi }{L},n=1,2,3,\dots$

• $n$ — count karta hai tube mein kitne half-waves fit hote hain. Yeh "neatly fit hona" condition hai.
• $A$ — sirf loudness hai; yeh drop out ho jaata hai kyunki equation kisi bhi amplitude par khush rehti hai.

---

Step 4 — Fitted shapes ko frequencies mein convert karo

KYA. Ab hum allowed $k_n$ jaante hain. Har ek ko ek actual pitch $f_n$ mein convert karo.

KYUN. $k$ waves per metre count karta hai; insaan aur hardware wobbles per second ki care karte hain. Bridge hai $\omega =ck$ (Step 2 se) aur $\omega =2\pi f$ (ek full circle $2\pi$ radians hai, aur $f$ full circles per second).

${\omega }_n=c{k}_n=\frac{n\pi c}{L},f_n=\frac{{\omega }_n}{2\pi }=\frac{n\pi c}{2\pi L}$

$\pi$'s cancel ho jaate hain, parent note ka headline result milta hai:

Link: yeh Standing waves and resonance result hot-gas tube par specialise kiya gaya hai.

---

Step 5 — Wahi idea, do extra directions (mode zoo)

KYA. Real chamber ek cylinder hai, 1-D line nahi. Waves across aur around bhi bounce kar sakti hain.

KYUN matter karta hai. Axial (length-wise) waves jo humne abhi banayi woh Longitudinal modes hain. Lekin sabse dangerous wali sideways slosh karti hain.

• Longitudinal (L): axis ke saath — $L$ se set, woh mode jise humne derive kiya.
• Tangential (T): axis ke around spinning — usually sabse zyada destructive, kyunki peak injector face ko sweep karta hai.
• Radial (R): centre line se andar-bahar saans lena.

Round cross-section ke liye "neatly fit hona" condition plain cosines ki jagah Bessel-function roots use karti hai:

$f_{mn}^{\text{trans}}=\frac{{\alpha }_{mn}c}{2\pi R}$

• $R$ — chamber radius.
• ${\alpha }_{mn}$ — $n\pi$ ka sideways version: woh special numbers jahan round drumhead ki wave rim par zero slope rakhti hai. Wahi logic (wall par zero velocity), curved geometry.

Tumhe yahan Bessel roots compute nahi karne — bas pattern jaano: walls zero velocity impose karti hain, sirf certain shapes fit hoti hain, har fitted shape ek mode hai.

---

Step 6 — Ek mode EXPLODE kyun kar sakta hai: swing aur flame

KYA. Ab fuel jalta hai. Burning heat $q^{\prime }$ dump karta hai (fluctuating heat-release). Hum poochte hain: kya woh heat wobble ko feed karti hai ya fight?

SWING analogy kyun exact hai. Swing tabhi height gain karti hai jab tum push karo jab woh tumhari push ki direction mein move kar rahi ho. Galat waqt push karo toh slow ho jaayegi. Acoustic wave ke liye, "pushing" matlab gas pehle se compressed hone par (pressure high) heat add karna. Squished gas mein add ki gayi heat use aur expand karti hai — woh direction jis mein woh pehle se ja rahi hai, usi mein ek shove.

Hum ek full cycle mein wave ko di gayi energy measure karte hain. Pressure $p^{\prime }\propto \cos (\omega t)$ lo aur heat $\varphi$ phase se lag kare: $q^{\prime }\propto \cos (\omega t-\varphi )$.

$\Delta E\propto {\int }_0^{2\pi /\omega }\underset{\text{pressure}}{\underbrace{\cos (\omega t)}}\underset{\text{heat}}{\underbrace{\cos (\omega t-\varphi )}}dt\propto \cos \varphi$

• Integral ${\int }_0^{2\pi /\omega }\dots dt$ — ek poore cycle mein push ka grand total add karta hai ($\int$ symbol sirf "grand total of" matlab hai).
• $\varphi$ — phase lag: pressure se kitne degrees peeche heat aati hai.
• Poori cheez $\frac{1}{2}\cos \varphi$ mein collapse ho jaati hai, kyunki $\varphi$ se alag do cosines ka average $\frac{1}{2}\cos \varphi$ hota hai.

$\cos \varphi$ dial padhna:

$\varphi$	$\cos \varphi$	Matlab
$0^{\circ }$	$+1$	heat exactly peak par → max drive, unstable
$90^{\circ }$	$0$	heat up-swing par → neutral
$180^{\circ }$	$-1$	heat trough par → damps, stable

---

Step 7 — Edge cases (reader ko kabhi stranded mat chhodho)

POORA STEP KYUN. Formula $f_n=nc/2L$ aur dial $\cos \varphi$ teen degenerate cases chhupa dete hain jo tumhe ZAROOR milenge.

• $n=0$? Koi mode nahi — $\cos (0)=1$ har jagah matlab uniform pressure, koi wobble nahi, zero frequency. Sabse chhota real mode $n=1$ hai.
• Cold-gas mistake. $c\approx 340$ m/s (room air) use karna $\sim 1000$ m/s (3000 K products) ki jagah har $f_n$ ko roughly $3\times$ underestimate karta hai. Hamesha hot $c=\sqrt{\gamma RT/M}$ use karo.
• $\varphi =90^{\circ }$ knife-edge. $\cos 90^{\circ }=0$: mode heat se na grow karta hai na decay — akele heat se. Choti extra losses phir ise stable tip karti hain; chhoti extra drive unstable tip karti hai — ek designer yahan kabhi bhi rehna nahi chahta.
• Amplitude ≠ danger. $\varphi =180^{\circ }$ par bada $q^{\prime }$ actively stabilise karta hai. $\cos \varphi$ ke galat sign ke saath badi heat tumhara dost hai.

---

Worked examples (pictures par re-derive kiye gaye)

---

Ek picture summary

Poori chain, left to right: wobble shoves neighbour ($p^{\prime \prime }$) → standing shape walls mein fit hoti hai ($k_n=n\pi /L$) → pitch ($f_n=nc/2L$) → flame ki timing ($\cos \varphi$) decide karta hai grow ya die.

Recall Feynman: poora walkthrough plain words mein

Gas ko ek jagah squeeze karo aur woh apne neighbour ko dhakelta hai — woh shove sound ki speed par travel karta hai, aur agar shape gas ko aur tezi se pile up karane ke liye bend kare, toh wahi wave equation hai. Resonance apni shape rakhti hai aur bas ek steady beat mein saans andar-bahar leti hai. Walls gas ko pass nahi hone deti, toh gas unke bilkul paas move nahi kar sakta — aur "wall par koi motion nahi" matlab pressure wahan sabse zyada bunched up hai, ek antinode. Sirf kuch cosine shapes ke flat tops bilkul dono ends par hote hain; wahi modes hain, aur har ek ko pitch mein convert karne par $f_n=nc/2L$ milta hai. Longer tube, lower hum; hotter gas, higher hum. Phir aag shamil hoti hai. Agar flame exactly tab sabse tej ho jab gas sabse zyada squeezed ho, toh yeh swing ko arc ke top par push karne jaisa hai — hum badhti jaati hai jab tak engine hil ke apart nahi ho jaata. Agar flame galat waqt par tej ho toh swing ko maar deta hai. Poori kahani ek dial hai: $\cos \varphi$, pressure ke against flame ki timing.

Recall Quick self-test

Sabse chhota real mode number? ::: $n=1$ (n=0 sirf uniform pressure hai, koi wobble nahi). Rigid wall par kya zero hai — pressure ya velocity? ::: Velocity; isse pressure ek antinode ban jaata hai. Grow vs. damp decide karne wali ek quantity kaunsi hai? ::: $\cos \varphi$, heat vs. pressure ka phase. Longer chamber → higher ya lower fundamental? ::: Lower ($f_n$ ke denominator mein L hai).