Gordon Nichols posted:That's what I found out with my recent carb "upgrade" - The carb venturii "choke point" was too big to match the rest of the engine flow so the entire system suffered. It's all supposed to be a matched system from air cleaner to exhaust tip and that sometimes takes a little figuring to make it all compatible because of all the variables.
An engine is an air pump, and power is based on how much air (and fuel) it will move through the engine. That's why cylinder heads are always advertised by CFM at various valve lifts.
A typical high-performance "streetable" camshaft will provide about .500 +/- of valve lift, so that's a good number to use for head capacity, and charts are always available to compare: generally using 25" of W/C as the basis for comparison, but some porters use 28". The formula to convert 25" ratings to 28" ratings is √(28 ÷ 25) or multiply the 25" rating by 1.058. This is a mathematical approximation, not an actual tested and published flow number, so it can only be used as a baseline.
What we don't generally know is how much air a typical Weber (or Dellorto) carb will flow, because Webers and Dellorotos (and Solexs, etc.) are rated by throttle plate size, as opposed to CFM like a Holly. Enter the interweb to provide some clues. A search resulted in the following information posted in a couple of different forums and cross-checked with each other:
- Dellorto 48 DRLA: 388 cfm/venturi at 28" W/C, or 358 cfm at 25" W/C
- Weber 48 IDF: 340 cfm/venturi at 28" W/C or, 313 cfm at 25" W/C
- Dellorto 45 DRLA: 329.3 cfm/venturi at 28" W/C, or 311.24 cfm at 25" W/C
- Weber 44 IDF: 290 cfm/venturi at 28" W/C, or 267 cfm at 25" W/C
- Dellorto 40 DRLA: 266 cfm/venturi at 28" W/C, or 245 cfm at 25" W/C
- Weber 40 IDF: 212 cfm/venturi at 28" W/C, or 195 cfm at 25" W/C
- Dellorto 36 DRLA: 206 cfm/venturi at 28" W/C, or 190 cfm at 25" W/C
All of this assumes stock venturis for the carbs in question. You can see by the chart that Dellortos flow about as much as the next size up Weber.
Anyhow, according to the current CB website, the following heads flow the following amount of air at .500 lift and 25" W/C:
- Panchitos: 167 cfm
- Wedge Ports: 201 cfm
- Comp Eliminator: 219 cfm
Based on this information, an engine with Wedge Ports should run Weber 40 IDFs. However, in real world dyno tests, more power is made with much bigger carbs (probably 48 DRLAs or at least 48 IDFs). Math is great, but real-world dyno numbers trump them every single time. What builders will tell you is that if the carb is even close to the choke-point, you're going to leave a lot of power on the table.
The venturis have to be small enough to increase airspeed enough to pick up the fuel through the jets well, but big enough to not restrict flow unduly. Carbs are (in my mind) a miracle of the mechanical age, at least on par with a mechanical watch. That they work at all is amazing, that they work well is astonishing. Why they are as big as they are on an air-cooled engine is something I couldn't just accept until very recently.
There's another thing that matters besides raw HP numbers on the dyno, and that is throttle response in the real world. Throttle response is a function of compression and airspeed velocity (which is measured in FPM). Higher compression ratios give better throttle response-- but there's a limit to what you can get away with, based on combustion efficiency and cooling. Bad cooling is baked in the air-cooled cake (which is why water-cooled engines have such good VEs), but a lot of strides have been made in combustion-chamber efficiency in the last 15 years.
Port velocity (FPM) information is never published, but we can make assumptions based on intake port volume (measured in CCs). This information is often published, and is pretty important. Generally speaking, when selecting heads, we are looking for the smallest intake port volume for a specific CFM rating. What we are looking for are head flow numbers that match the target HP we are looking for (there's a formula for this as well), then looking for the smallest ports available that will flow that number. One has to be careful, as hand porting varies by the nature of being done by hand. As such, I (and the tuner community generally) have become a fan of CNC (or "as cast") ports, which are consistent across individual ports.
This stuff matters in how an engine feels. It's easy to get totally lost in the math and formulas, and to start to think that smart guys who've made their lives modifying and building engines are missing something. This is hogwash. At some point, you've got to let go of the tables and multipliers, and just go with what somebody you trust says will work.
That's super-hard for a hard-headed MKA (Mr. Know-it-All) like your's truly.