Openai/6902a8c8-76f4-8001-a86a-93fe7b5e36c6


This conversation was exported from ChatGPT[1]

gpt-5-t-mini edit

User: I want to do some final checks on my intake. edit

I want to do some final checks on my intake.

  1. The length of each runner is 576.6mm or 22.7". Each runners inner diameter is 41.148mm and has no tapper. 190mm from the cylinder head their is a 180° bend which puts the centerline of each end of the u pipe 6" away from each other.
  2. Using the D shaped plenum that is 4" by 4" that I uploaded a picture of earlier, it measures 12" long.
  3. The throttle body will be the jeep 4.0 throttle body with an opening diameter of 56mm.
  4. I will be adding long tube headers to the engine as well as removing the cat for more flow.

considering all these variables how does my truck stand to perform? Will it meet the targets I had described earlier? How will the idle perform?

Assistant: Nice — thanks, that’s enough to give a concrete assessment. edit

Nice — thanks, that’s enough to give a concrete assessment. I ran the numbers and put together a practical summary plus a few small recommendations.

Quick numbers (converted) edit

  • Runner length: 576.6 mm (22.7")
  • Runner ID: 41.148 mm (≈1.62") → single-runner cross-sectional area ≈ 0.00133 m²
  • Throttle body: 56 mm → TB area ≈ 0.00246 m²
  • Plenum (D-shape 4"×4"×12"): 192 in³ ≈ 3,146 cc ≈ 3.15 L total (≈0.79 L per cylinder)
  • Engine assumed: 22R ≈ 2.366 L (2366 cc), 4-cyl (used for flow examples)

Flow / velocity check (useful for idle & drivability) edit

Using engine displacement to get approximate bulk airflow, the per-runner air velocity estimates are:

  • Idle (~800 rpm): ~3.0 m/s in each runner
  • 1000 rpm: ~3.7 m/s
  • 2000 rpm: ~7.4 m/s
  • 3000 rpm: ~11.1 m/s

Through the throttle bore the air velocity is noticeably higher (because TB area < total runner area). at 3000 rpm TB ≈ 24 m/s. That means the TB will be the higher-velocity (and potential) restriction point compared to a single runner, which is normal.

What this geometry will give you (practical, non-mathematical) edit

  • Strong low-end and midrange torque / very responsive low RPM: Long runners (22.7") with a relatively small runner diameter (41 mm) strongly favor torque and cylinder filling at low–mid RPM. That’s the classic recipe for street torque and good drivability.
  • Good idle stability expected: The combination of long, relatively narrow runners and a modest plenum (3.15 L total / ~0.79 L per cylinder) promotes good air velocity at idle which helps atomization and stable idle—provided the throttle body / idle control is set up properly.
  • Throttle body sizing: The 56 mm TB has less cross-section than the sum of the runners (normal). That makes the TB the higher-velocity throat. For everyday street use and a torque-focused tune this TB is fine. At very high RPM it could be a limiting restriction vs. a larger TB, but for a 22R used on the street it isn’t unusually small.
  • Plenum volume effects: ~3.15 L total is a relatively small plenum (helps responsiveness and midrange). It supports your torque-oriented goal; you won’t get the “large-plenum high-RPM” behaviour that favors peak HP above where the runners tune.
  • Headers & cat delete: Long-tube headers and removing the cat reduce backpressure and will help top-end and midrange power. Combined with your long runners they’ll likely broaden the torque band and maybe raise VE across the midrange — a good match for your intent.

Potential downsides / caveats edit

  • High-RPM peak horsepower tradeoff: Very long runners push the tuned-wave benefits down in RPM. If your original targets were high peak HP at high RPM (e.g., >5–6k RPM peak), this runner length will make it harder to reach that peak. If your targets were low-to-mid RPM torque (street torque, towing, rolling acceleration), you’re well poised to hit them.
  • 180° U-bend location & sharpness: a tight 180° bend can create flow separation and turbulence. Make that bend as smooth as practical, with generous radius and well blended transitions into the runner to avoid losses. Sharp edges at the head/plenum interface will hurt VE more than you’d expect.
  • Throttle control & idle tuning: a 56 mm TB plus long runners can produce a lopey idle if idle air control (IAC) or ECU fueling/idle map isn’t tuned. If running an idle valve or ISC, make sure it’s functional and the idle stop is set for a predictable position. If carbureted, you may need smaller bores or a good jetting/tuning session.
  • Port matching: ensure the runner ID and the head port are matched (or properly tapered). Sudden contraction/expansion at the head undermines the runner tuning benefit.
  • Resonance details: I avoided giving a single “tuned-RPM” number because intake wave reflection timing depends on valve timing, effective length (including head trumpets, corrections), and temperature. Qualitatively — this length favors idle→midrange tuning, not a high-RPM quarter-wave tune.

Practical recommendations (what to check/do next) edit

  1. Smooth the U-bend — radius ≥ ~2× runner diameter if possible and blend with the rest of the runner.
  2. Match runner exits to the head port (and use short, smooth velocity stacks/trumpets at the head entrance if you can). Even small mismatches cost VE.
  3. Ensure plenum runout is smooth and the throttle plate is centered and free of abrupt step changes. Consider a small velocity step or short entrance trumpet to help distribution if you see uneven fueling.
  4. Idle control / tune — make sure idle servo/IAC and fueling are tuned after fitment; expect to retune base fueling and ignition timing after manifold/header/cat changes.
  5. If you want more high-RPM power: shorten the runners (or add trumpet length tuning adapters that are removable), and/or increase plenum volume. If you want to keep the long-runner character but shift power higher, shorten a bit (even a few inches) helps a lot.
  6. Dyno or data log: final confirmation should be on a dyno or by controlled runs with AFR/logging. The geometry gives a very good prediction, but real VE/flow and transient effects are what determine final numbers.

Short verdict edit

  • If your targets were low-end torque, drivability, and strong midrange: this design is well-matched and should meet or exceed those expectations (assuming proper tuning).
  • If your targets were high-RPM peak horsepower: this runner length and plenum will likely push your peak down; you may not hit a high-RPM horsepower target without shortening runners or changing plenum/venturi sizing.

If you want, I can:

  • Convert the geometry into an estimated tuned band (give a rough tuned RPM band) — I can do that with the usual quarter-wave approximations and state the assumptions; or
  • Run a quick “what if” (shorten runners by X inches or change TB to Y mm) so you can see how the band and velocity numbers change.

Which would you prefer?

References edit

  1. "Performance assessment and idle". ChatGPT. Retrieved 2026-01-14.